Compare commits

..

12 Commits

Author SHA1 Message Date
KyuubiYoru 9e863ebf64 docs(integration): verify Godot and Linux SpaceGame pilot (#21)
quality-gate / quality (push) Failing after 1m40s
quality-gate / container (push) Has been skipped
2026-07-16 20:32:53 +02:00
KyuubiYoru ebb5eb617c docs(integration): record SpaceGame pilot checkpoint (#21)
quality-gate / quality (push) Failing after 1m26s
quality-gate / container (push) Has been skipped
2026-07-16 19:18:03 +02:00
KyuubiYoru 7fb85059fb docs: add integration guides and incident runbooks (#20)
quality-gate / quality (push) Failing after 1m28s
quality-gate / container (push) Has been skipped
2026-07-16 18:25:10 +02:00
KyuubiYoru cc5793f935 feat(release): add reproducible signed artifacts (#19)
quality-gate / quality (push) Failing after 1m50s
quality-gate / container (push) Has been skipped
2026-07-16 17:48:21 +02:00
KyuubiYoru 07004cd75f docs(evidence): record clean capacity candidate (#18)
quality-gate / quality (push) Failing after 1m28s
quality-gate / container (push) Has been skipped
2026-07-16 16:11:39 +02:00
KyuubiYoru cf14836d48 fix(operations): require clean candidate provenance (#18) 2026-07-16 16:04:25 +02:00
KyuubiYoru 609dad7cf1 feat(operations): add capacity and resilience gates (#18) 2026-07-16 15:57:01 +02:00
KyuubiYoru 08729ae25c feat(deployment): add secure Linux runtime (#17)
quality-gate / quality (push) Failing after 1m9s
quality-gate / container (push) Has been skipped
2026-07-16 15:03:04 +02:00
KyuubiYoru be732de7c9 feat(server): add observability and operator controls (#16)
quality-gate / quality (push) Failing after 1m1s
2026-07-16 13:22:16 +02:00
KyuubiYoru 88ef946af5 feat(server): harden hostile input and overload behavior (#15)
quality-gate / quality (push) Failing after 1m5s
2026-07-16 12:37:38 +02:00
KyuubiYoru 2ff7cd6d9d test(integration): add deterministic NAT topology harness (#14)
quality-gate / quality (push) Failing after 1m3s
2026-07-16 11:50:53 +02:00
KyuubiYoru 7e3be2cad1 feat(tooling): add standalone rendezvous test client (#25)
quality-gate / quality (push) Failing after 1m3s
2026-07-16 11:05:56 +02:00
130 changed files with 17764 additions and 320 deletions
+13
View File
@@ -0,0 +1,13 @@
.git
.gitea
.idea
.vs
.codex
.agents
**/bin
**/obj
TestResults
deploy/compose/secrets
deploy/compose/.smoke.env
docs
tests
+136 -2
View File
@@ -14,16 +14,34 @@ jobs:
timeout-minutes: 15
steps:
- name: Check out repository
uses: actions/checkout@v4
uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
with:
fetch-depth: 0
- name: Install .NET SDK
uses: actions/setup-dotnet@v4
uses: actions/setup-dotnet@67a3573c9a986a3f9c594539f4ab511d57bb3ce9 # v4.3.1
with:
dotnet-version: 10.0.301
- name: Restore locked dependencies
run: dotnet restore Rendezvous.slnx --locked-mode
- name: Verify dependency licenses and reviewed transport pin
run: python3 eng/release_artifacts.py policy --root .
- name: Reject vulnerable direct or transitive packages
shell: bash
run: |
set -euo pipefail
dotnet package list --project Rendezvous.slnx \
--vulnerable --include-transitive --no-restore --format json \
>"${RUNNER_TEMP}/nuget-vulnerabilities.json"
python3 eng/release_artifacts.py audit \
--input "${RUNNER_TEMP}/nuget-vulnerabilities.json"
- name: Enforce compatibility version bumps
run: ./scripts/check-compatibility.sh origin/main
- name: Verify formatting and analyzers
run: dotnet format Rendezvous.slnx --verify-no-changes --no-restore
@@ -35,3 +53,119 @@ jobs:
- name: Test
run: dotnet test Rendezvous.slnx --configuration Release --no-build
- name: Run quick capacity and resilience gate
run: ./scripts/run-capacity-gate.sh
- name: Test privileged Linux namespace topology when available
shell: bash
run: |
set -euo pipefail
probe="rendezvous-probe-$$"
suffix="$(( $$ % 100000 ))"
bridge="rvb${suffix}"
veth_root="rvr${suffix}"
veth_peer="rvp${suffix}"
cleanup_probe() {
if [[ -n "$veth_root" ]]; then
ip link delete "$veth_root" >/dev/null 2>&1 || true
fi
if [[ -n "$bridge" ]]; then
ip link delete "$bridge" >/dev/null 2>&1 || true
fi
if [[ -n "$probe" ]]; then
ip netns delete "$probe" >/dev/null 2>&1 || true
fi
}
trap cleanup_probe EXIT
if command -v ip >/dev/null 2>&1 \
&& command -v iptables >/dev/null 2>&1 \
&& command -v sysctl >/dev/null 2>&1 \
&& ip netns add "$probe" 2>/dev/null \
&& ip link add "$bridge" type bridge \
&& ip link add "$veth_root" type veth peer name "$veth_peer" \
&& ip link set "$veth_root" master "$bridge" \
&& ip link set "$veth_peer" netns "$probe" \
&& ip netns exec "$probe" sysctl -q -w net.ipv4.ip_forward=1 \
&& ip netns exec "$probe" iptables -t nat -A POSTROUTING -o "$veth_peer" -j MASQUERADE \
&& ip netns exec "$probe" iptables -A FORWARD -i "$veth_peer" -o lo \
-m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT; then
ip link delete "$veth_root"
veth_root=""
ip link delete "$bridge"
bridge=""
ip netns delete "$probe"
probe=""
results="${RUNNER_TEMP:-/tmp}/rendezvous-netns-results"
mkdir -p "$results"
RENDEZVOUS_RUN_NETNS_TESTS=1 dotnet test Rendezvous.slnx \
--configuration Release \
--no-build \
--filter FullyQualifiedName~PrivilegedLinuxNatNamespacesCompleteDirectTrafficAcrossSeparateObservedEndpoints \
--logger "trx;LogFileName=netns.trx" \
--results-directory "$results"
grep -q 'testName="[^"]*\.PrivilegedLinuxNatNamespacesCompleteDirectTrafficAcrossSeparateObservedEndpoints"' \
"$results/netns.trx"
else
echo "Network namespaces/NAT tooling unavailable; deterministic loopback topology remains the required gate."
fi
container:
needs: quality
runs-on: ubuntu-latest
timeout-minutes: 15
steps:
- name: Check out repository
uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Install .NET SDK
uses: actions/setup-dotnet@67a3573c9a986a3f9c594539f4ab511d57bb3ce9 # v4.3.1
with:
dotnet-version: 10.0.301
- name: Build deployment diagnostic
run: |
dotnet restore src/FinalFactory.Rendezvous.TestClient/FinalFactory.Rendezvous.TestClient.csproj --locked-mode
dotnet build src/FinalFactory.Rendezvous.TestClient/FinalFactory.Rendezvous.TestClient.csproj --configuration Release --no-restore
- name: Build and exercise hardened container
shell: bash
run: |
set -euo pipefail
compose_file="deploy/compose/compose.yaml"
secret="deploy/compose/secrets/signing-key"
cleanup() {
RENDEZVOUS_UID=1654 RENDEZVOUS_GID=1654 \
docker compose -f "$compose_file" down --volumes >/dev/null 2>&1 || true
rm -f "$secret"
}
trap cleanup EXIT
install -d -m 0700 deploy/compose/secrets
openssl rand -out "$secret" 32
chmod 0444 "$secret"
export RENDEZVOUS_UID=1654
export RENDEZVOUS_GID=1654
export SOURCE_REVISION_ID="$GITHUB_SHA"
docker compose -f "$compose_file" build \
--build-arg SOURCE_REVISION_ID="$SOURCE_REVISION_ID"
docker compose -f "$compose_file" up --no-build --detach
container_id="$(docker compose -f "$compose_file" ps -q rendezvous)"
test -n "$container_id"
test "$(docker inspect --format '{{.Config.User}}' "$container_id")" = "1654:1654"
test "$(docker inspect --format '{{.HostConfig.ReadonlyRootfs}}' "$container_id")" = "true"
test "$(docker inspect --format '{{range .Mounts}}{{if eq .Destination \"/app/appsettings.Production.json\"}}{{.RW}}{{end}}{{end}}' "$container_id")" = "false"
test "$(docker inspect --format '{{range .Mounts}}{{if eq .Destination \"/run/secrets/rendezvous-signing-key\"}}{{.RW}}{{end}}{{end}}' "$container_id")" = "false"
for attempt in {1..100}; do
if curl --fail --silent http://127.0.0.1:8080/health/ready >/dev/null 2>&1; then
break
fi
if (( attempt == 100 )); then
docker compose -f "$compose_file" logs rendezvous
exit 1
fi
sleep 0.1
done
./scripts/smoke-deployment.sh
docker compose -f "$compose_file" stop --timeout 40 rendezvous
test "$(docker inspect --format '{{.State.Running}}' "$container_id")" = "false"
test "$(docker inspect --format '{{.State.ExitCode}}' "$container_id")" = "0"
+193
View File
@@ -0,0 +1,193 @@
name: immutable-release
on:
push:
tags:
- "v*.*.*"
concurrency:
group: release-${{ gitea.ref_name }}
cancel-in-progress: false
jobs:
release:
runs-on: ubuntu-latest
timeout-minutes: 45
environment: production
steps:
- name: Check out immutable tag
uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
with:
fetch-depth: 0
- name: Install pinned .NET SDK
uses: actions/setup-dotnet@67a3573c9a986a3f9c594539f4ab511d57bb3ce9 # v4.3.1
with:
dotnet-version: 10.0.301
- name: Install pinned Buildx and BuildKit
uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3.11.1
with:
version: v0.35.0
install: true
driver-opts: image=moby/buildkit:v0.25.2@sha256:0f63d66f8d2de0bd16438284831a3e9ee6ca7cd57b6eb3ed6e38a7a456590fa7
- name: Validate tag and produce reproducible artifacts
shell: bash
run: |
set -euo pipefail
version="${GITHUB_REF_NAME#v}"
./scripts/check-release-tag.sh "$GITHUB_REF_NAME"
previous_tag="$(git tag --merged HEAD^ --list 'v*.*.*' --sort=-version:refname | sed -n '1p')"
if [[ -n "$previous_tag" ]]; then
./scripts/check-compatibility.sh "$previous_tag"
elif [[ -n "$(git tag --list 'v*.*.*' | sed -n '1p')" ]]; then
echo "No prior release tag is an ancestor of $GITHUB_REF_NAME." >&2
exit 1
else
./scripts/check-compatibility.sh __initial_release_without_base__
fi
release_builder="rendezvous-release-builder:${GITHUB_SHA}"
docker buildx build \
--platform linux/amd64 \
--file eng/release-builder.Dockerfile \
--target release-builder \
--load \
--tag "$release_builder" .
mkdir -p "${RUNNER_TEMP}/release-home" "${RUNNER_TEMP}/nuget"
docker run --rm \
--user "$(id -u):$(id -g)" \
--env HOME="${RUNNER_TEMP}/release-home" \
--env NUGET_PACKAGES="${RUNNER_TEMP}/nuget" \
--volume "$GITHUB_WORKSPACE:/source" \
--volume "${RUNNER_TEMP}:${RUNNER_TEMP}" \
--workdir /source \
"$release_builder" \
./scripts/build-release.sh "$version" "${RUNNER_TEMP}/release/$version"
./scripts/verify-real-consumers.sh "$version" "${RUNNER_TEMP}/release/$version"
echo "RENDEZVOUS_VERSION=$version" >>"$GITHUB_ENV"
echo "RENDEZVOUS_RELEASE_DIR=${RUNNER_TEMP}/release/$version" >>"$GITHUB_ENV"
echo "RENDEZVOUS_RELEASE_BUILDER=$release_builder" >>"$GITHUB_ENV"
- name: Build exact container candidate
shell: bash
run: |
set -euo pipefail
export SOURCE_DATE_EPOCH="$(git show -s --format=%ct HEAD)"
common=(
--no-cache
--pull=false
--provenance=false
--platform linux/amd64
--build-arg SOURCE_DATE_EPOCH="$SOURCE_DATE_EPOCH"
--build-arg SOURCE_REVISION_ID="$GITHUB_SHA"
)
release_tag="git.finalfactory.de/heikyu/rendezvous:${RENDEZVOUS_VERSION}"
image_one="${RUNNER_TEMP}/rendezvous-image-1.tar"
image_two="${RUNNER_TEMP}/rendezvous-image-2.tar"
docker buildx build "${common[@]}" --tag "$release_tag" \
--output "type=docker,dest=$image_one,rewrite-timestamp=true" .
docker buildx build "${common[@]}" --tag "$release_tag" \
--output "type=docker,dest=$image_two,rewrite-timestamp=true" .
cmp --silent "$image_one" "$image_two"
docker load --input "$image_one"
candidate_id="$(docker image inspect --format '{{.Id}}' "$release_tag")"
buildkit_version="$(docker buildx inspect --bootstrap | sed -n 's/.*BuildKit version: *//p' | sed -n '1p')"
docker run --rm \
--user "$(id -u):$(id -g)" \
--volume "$GITHUB_WORKSPACE:/source" \
--volume "${RUNNER_TEMP}:${RUNNER_TEMP}" \
--workdir /source \
"$RENDEZVOUS_RELEASE_BUILDER" \
python3 eng/release_artifacts.py record-container-build \
--provenance "$RENDEZVOUS_RELEASE_DIR/release-provenance.json" \
--buildx-version "$(docker buildx version)" \
--buildkit-version "$buildkit_version" \
--image-id "$candidate_id"
- name: Stage HTTP registration, browse, and authenticated UDP traversal
shell: bash
run: |
set -euo pipefail
secret="deploy/compose/secrets/signing-key"
cleanup() {
RENDEZVOUS_UID=1654 RENDEZVOUS_GID=1654 RENDEZVOUS_IMAGE="git.finalfactory.de/heikyu/rendezvous:${RENDEZVOUS_VERSION}" \
docker compose -f deploy/compose/compose.yaml down --volumes >/dev/null 2>&1 || true
rm -f "$secret"
}
trap cleanup EXIT
install -d -m 0700 deploy/compose/secrets
openssl rand -out "$secret" 32
chmod 0444 "$secret"
export RENDEZVOUS_UID=1654
export RENDEZVOUS_GID=1654
export RENDEZVOUS_IMAGE="git.finalfactory.de/heikyu/rendezvous:${RENDEZVOUS_VERSION}"
docker compose -f deploy/compose/compose.yaml up --detach --no-build
for attempt in {1..100}; do
curl --fail --silent http://127.0.0.1:8080/health/ready >/dev/null 2>&1 && break
if (( attempt == 100 )); then
docker compose -f deploy/compose/compose.yaml logs rendezvous
exit 1
fi
sleep 0.1
done
./scripts/smoke-deployment.sh
- name: Scan candidate for high and critical vulnerabilities
uses: aquasecurity/trivy-action@57a97c7e7821a5776cebc9bb87c984fa69cba8f1 # v0.35.0, post-incident safe SHA
with:
image-ref: git.finalfactory.de/heikyu/rendezvous:${{ env.RENDEZVOUS_VERSION }}
version: v0.69.3
format: table
exit-code: "1"
ignore-unfixed: false
severity: HIGH,CRITICAL
- name: Generate container SPDX inventory
uses: aquasecurity/trivy-action@57a97c7e7821a5776cebc9bb87c984fa69cba8f1 # v0.35.0, post-incident safe SHA
with:
image-ref: git.finalfactory.de/heikyu/rendezvous:${{ env.RENDEZVOUS_VERSION }}
version: v0.69.3
format: spdx-json
output: ${{ env.RENDEZVOUS_RELEASE_DIR }}/FinalFactory.Rendezvous.Container.${{ env.RENDEZVOUS_VERSION }}.spdx.json
- name: Finalize checksums over the publish-ready candidate
shell: bash
run: |
set -euo pipefail
source_date_epoch="$(git show -s --format=%ct HEAD)"
docker run --rm \
--user "$(id -u):$(id -g)" \
--volume "$GITHUB_WORKSPACE:/source" \
--volume "${RUNNER_TEMP}:${RUNNER_TEMP}" \
--workdir /source \
"$RENDEZVOUS_RELEASE_BUILDER" \
bash -c 'python3 eng/release_artifacts.py normalize-container-sbom \
--file "$1/FinalFactory.Rendezvous.Container.$2.spdx.json" \
--version "$2" \
--commit "$3" \
--source-date-epoch "$4" \
&& ./scripts/finalize-release-candidate.sh "$2" "$1"' \
_ "$RENDEZVOUS_RELEASE_DIR" "$RENDEZVOUS_VERSION" "$GITHUB_SHA" "$source_date_epoch"
- name: Preserve verified candidate artifacts
uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
with:
name: rendezvous-${{ env.RENDEZVOUS_VERSION }}
path: ${{ env.RENDEZVOUS_RELEASE_DIR }}
if-no-files-found: error
retention-days: 30
- name: Install pinned signing client
uses: sigstore/cosign-installer@6f9f17788090df1f26f669e9d70d6ae9567deba6 # v4.1.2
with:
cosign-release: v3.0.6
- name: Publish once, sign, attest, and create release
shell: bash
env:
RENDEZVOUS_RELEASE_USERNAME: ${{ secrets.RELEASE_USERNAME }}
RENDEZVOUS_RELEASE_TOKEN: ${{ secrets.RELEASE_TOKEN }}
COSIGN_PRIVATE_KEY: ${{ secrets.COSIGN_PRIVATE_KEY }}
COSIGN_PASSWORD: ${{ secrets.COSIGN_PASSWORD }}
run: ./scripts/publish-release.sh "$RENDEZVOUS_VERSION" "$RENDEZVOUS_RELEASE_DIR"
+6
View File
@@ -6,3 +6,9 @@ TestResults/
*.suo
*.user
*.userosscache
deploy/compose/.smoke.env
artifacts/
__pycache__/
*.pyc
deploy/compose/secrets/*
!deploy/compose/secrets/.gitignore
+25
View File
@@ -0,0 +1,25 @@
# Changelog
All notable Rendezvous release changes are recorded here. Versions follow
Semantic Versioning; HTTP, UDP, and connection-ticket format compatibility is
tracked separately and called out for every release.
## 1.0.0 - 2026-07-16
### Compatibility
- Initial Client and Contracts package major: 1.
- HTTP contract: v1; UDP mediation contract: v1; connection-ticket format: v1.
- Server accepts Client 1.0.0 through the latest compatible 1.x release.
- Client traversal is pinned to LiteNetLib 2.1.4; LiteNetLib 1.x is unsupported.
### Security and configuration
- Packages contain no reusable credentials or environment configuration.
- Production server startup requires provisioned signing keys and the hardened
single-active deployment configuration.
### Migration
- This is the first packaged release; no prior package or wire migration exists.
- Consumers must pin both Rendezvous packages to the same exact version.
+28
View File
@@ -1,4 +1,5 @@
<Project>
<Import Project="eng/Versions.props" />
<PropertyGroup>
<AnalysisLevel>latest-recommended</AnalysisLevel>
<ContinuousIntegrationBuild Condition="'$(CI)' == 'true'">true</ContinuousIntegrationBuild>
@@ -8,8 +9,35 @@
<ImplicitUsings>enable</ImplicitUsings>
<LangVersion>latest</LangVersion>
<Nullable>enable</Nullable>
<Version>$(RendezvousVersion)</Version>
<PackageVersion Condition="'$(PackageVersion)' == ''">$(RendezvousVersion)</PackageVersion>
<AssemblyVersion>$(RendezvousMajorVersion).0.0.0</AssemblyVersion>
<FileVersion>$(RendezvousMajorVersion).$(RendezvousMinorVersion).$(RendezvousPatchVersion).0</FileVersion>
<Authors>Final Factory</Authors>
<Company>Final Factory</Company>
<RepositoryUrl>https://git.finalfactory.de/HeiKyu/Rendezvous</RepositoryUrl>
<RepositoryType>git</RepositoryType>
<PackageProjectUrl>https://git.finalfactory.de/HeiKyu/Rendezvous</PackageProjectUrl>
<PublishRepositoryUrl>true</PublishRepositoryUrl>
<EmbedUntrackedSources>true</EmbedUntrackedSources>
<EnableSourceLink>true</EnableSourceLink>
<IncludeSymbols>true</IncludeSymbols>
<SymbolPackageFormat>snupkg</SymbolPackageFormat>
<PackageReleaseNotes>See CHANGELOG.md in the package and repository.</PackageReleaseNotes>
<RestorePackagesWithLockFile>true</RestorePackagesWithLockFile>
<RestoreLockedMode Condition="'$(CI)' == 'true'">true</RestoreLockedMode>
<NuGetAudit>true</NuGetAudit>
<NuGetAuditMode>all</NuGetAuditMode>
<NuGetAuditLevel>moderate</NuGetAuditLevel>
<TreatWarningsAsErrors>true</TreatWarningsAsErrors>
</PropertyGroup>
<Target Name="ConfigureGiteaSourceLink"
BeforeTargets="_GenerateSourceLinkFile"
DependsOnTargets="InitializeSourceControlInformation">
<ItemGroup>
<SourceRoot Update="@(SourceRoot)"
SourceLinkUrl="$(RepositoryUrl)/raw/commit/$(SourceRevisionId)/*" />
</ItemGroup>
</Target>
</Project>
+1 -1
View File
@@ -4,7 +4,7 @@
<CentralPackageTransitivePinningEnabled>true</CentralPackageTransitivePinningEnabled>
</PropertyGroup>
<ItemGroup>
<PackageVersion Include="LiteNetLib" Version="2.1.4" />
<PackageVersion Include="LiteNetLib" Version="[$(LiteNetLibVersion)]" />
<PackageVersion Include="Microsoft.AspNetCore.OpenApi" Version="10.0.9" />
<PackageVersion Include="Microsoft.Extensions.ApiDescription.Server" Version="10.0.9" />
<PackageVersion Include="Microsoft.NET.Test.Sdk" Version="18.4.0" />
+35
View File
@@ -0,0 +1,35 @@
# syntax=docker/dockerfile:1.7@sha256:a57df69d0ea827fb7266491f2813635de6f17269be881f696fbfdf2d83dda33e
FROM mcr.microsoft.com/dotnet/sdk:10.0.301-noble@sha256:ea8bde36c11b6e7eec2656d0e59101d4462f6bd630730f2c8201ed0572b295d5 AS build
ARG SOURCE_REVISION_ID
WORKDIR /source
COPY Directory.Build.props Directory.Packages.props NuGet.config global.json Rendezvous.slnx ./
COPY eng/Versions.props eng/Versions.props
COPY src/FinalFactory.Rendezvous.Contracts/FinalFactory.Rendezvous.Contracts.csproj src/FinalFactory.Rendezvous.Contracts/packages.lock.json src/FinalFactory.Rendezvous.Contracts/
COPY src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj src/FinalFactory.Rendezvous.Server/packages.lock.json src/FinalFactory.Rendezvous.Server/
RUN dotnet restore src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj --locked-mode
COPY src/FinalFactory.Rendezvous.Contracts/ src/FinalFactory.Rendezvous.Contracts/
COPY src/FinalFactory.Rendezvous.Server/ src/FinalFactory.Rendezvous.Server/
RUN dotnet publish src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj \
--configuration Release \
--no-restore \
--output /out \
/p:UseAppHost=false \
/p:OpenApiGenerateDocuments=false \
/p:ContinuousIntegrationBuild=true \
/p:RepositoryCommit="$SOURCE_REVISION_ID" \
/p:SourceRevisionId="$SOURCE_REVISION_ID"
FROM mcr.microsoft.com/dotnet/aspnet:10.0.9-noble-chiseled@sha256:f820c4fbfb8bb204c3bbe05c69d48cd039cd0e67aa8f13ac1cec168819b90643 AS runtime
ENV ASPNETCORE_HTTP_PORTS=8080 \
DOTNET_EnableDiagnostics=0 \
DOTNET_CLI_TELEMETRY_OPTOUT=1 \
TMPDIR=/tmp
WORKDIR /app
COPY --from=build --chown=1654:1654 /out/ ./
USER 1654:1654
EXPOSE 8080/tcp
EXPOSE 9050/udp
ENTRYPOINT ["dotnet", "FinalFactory.Rendezvous.Server.dll"]
+43 -6
View File
@@ -38,7 +38,11 @@ UDP hole punching cannot guarantee a direct connection through every network. Sy
- `FinalFactory.Rendezvous.TestClient` — thin interactive and scriptable host/browser/join diagnostic built only on the public SDK.
- `FinalFactory.Rendezvous.Tests` — unit, integration, security, and connection-lifecycle tests.
The server directory and NAT mediator begin as separate modules in one deployable service because they share session, lease, authorization, and endpoint state. Their internal boundary should allow independent deployment later if scale, availability, or security requirements diverge.
The server directory and NAT mediator are separate modules in one single-active
deployable service because they share ephemeral session, lease, authorization,
replay, and endpoint state. Their internal boundary can support a future
explicitly designed shared-state architecture; operators must not create
multiple active v1 replicas.
## Service boundaries
@@ -77,8 +81,11 @@ The initial service does not provide:
Rendezvous is under active roadmap development. The versioned contracts,
directory leases, authenticated join attempts, LiteNetLib mediator, caller-owned
SDK coordination, and typed connection outcomes are implemented. The thin test
client, deployment hardening, and production-readiness roadmap remain in progress;
SDK coordination, typed connection outcomes, thin public-SDK diagnostic client,
deterministic NAT topology harness, hostile-input controls,
observability/operator surface, secure single-active Linux deployment, and
numeric capacity/resilience gates, and reproducible signed release pipeline are
implemented. Consumer pilots and final production-readiness gates remain in progress;
participating games must not treat the current repository as a finished production
service until those gates land.
@@ -88,6 +95,34 @@ The frozen v1 wire surface is documented in the
[HTTP, UDP, and generated OpenAPI contracts](docs/contracts/README.md).
Tenant policy, publisher/operator principals, and production key custody are
defined in [game provisioning and signing-key lifecycle](docs/security/provisioning.md).
Layered HTTP/UDP budgets, overload behavior, and safe operational tuning are
defined in [hostile-input and overload protection](docs/security/abuse-protection.md).
Health semantics, bounded telemetry, alerting, audit privacy, and the authenticated
operator controls are defined in the
[observability and operator runbook](docs/operations/observability-and-operator-runbook.md).
Concrete detect/contain/recover/verify procedures are in the
[incident and change runbooks](docs/operations/incident-runbooks.md).
The pinned non-root container, production topology, graceful drain, Linux
hardening, smoke procedure, and recovery lifecycle are documented in
[secure single-active Linux deployment](docs/deployment/linux.md).
The numeric core-state candidate profile, public launch objectives, accelerated
soak, resilience matrix, and single-active scaling decision are recorded in
[capacity and resilience gates](docs/operations/capacity-and-resilience.md).
Release versions, compatibility windows, immutable artifact construction,
signing, staged promotion, rollback, and migration are defined in
[releases and compatibility](docs/releases/README.md).
The scriptable host/browser/join diagnostic and its stable automation contract are
documented in the [TestClient integration guide](docs/integration/test-client.md).
The package, gameplay-socket, host-admission, provisioning, metadata, key rotation,
versioning, and secure rollout seams are in the
[game integration guide](docs/integration/sdk-seams.md).
The always-on three-party scenarios, optional Linux namespace topology, and
simulation limits are documented in the
[deterministic topology harness](docs/integration/topology-harness.md).
The current consumer evidence and the still-open public-package, Godot-process,
fallback, reconnect, Linux, and external-NAT gates are tracked in the
[SpaceGame consumer pilot](docs/integration/spacegame-pilot.md).
## Development
@@ -104,9 +139,11 @@ dotnet test Rendezvous.slnx --configuration Release --no-build
Run the bootstrap server with
`dotnet run --project src/FinalFactory.Rendezvous.Server`. It serves HTTP health endpoints and binds
the configured UDP mediator port; both stop through normal host cancellation.
The launch profile uses an ephemeral development-only signing key. Production
startup fails closed until externally supplied game policies and `env:` signing
key references resolve to valid key material; no reusable game secret is stored
The launch profile uses separate ephemeral development-only publisher and operator
signing keys. Production
startup fails closed until its advertised endpoints, proxy trust boundary,
externally supplied game policies, and `env:` (base64) or `file:` (raw,
absolute, non-symlink) signing-key references resolve safely; no reusable game secret is stored
in this repository or the public Client package.
The project dependency rules and supported runtime choices are documented in
[project and dependency boundaries](docs/architecture/project-boundaries.md).
+1
View File
@@ -6,6 +6,7 @@
<Project Path="src/FinalFactory.Rendezvous.TestClient/FinalFactory.Rendezvous.TestClient.csproj" />
</Folder>
<Folder Name="/tests/">
<Project Path="tests/FinalFactory.Rendezvous.Capacity/FinalFactory.Rendezvous.Capacity.csproj" />
<Project Path="tests/FinalFactory.Rendezvous.Tests/FinalFactory.Rendezvous.Tests.csproj" />
</Folder>
</Solution>
@@ -0,0 +1,60 @@
{
"AllowedHosts": "localhost;127.0.0.1;rendezvous",
"Rendezvous": {
"Deployment": {
"PublicHttpBaseUrl": "https://localhost/",
"PublicUdpHost": "127.0.0.1",
"PublicUdpPort": 9050,
"DrainDeadlineSeconds": 30,
"MinimumDrainSeconds": 1,
"SingleActiveInstance": true,
"AllowPrivatePublicEndpoints": true
},
"Udp": {
"ListenAddress": "0.0.0.0",
"Port": 9050
},
"AbuseProtection": {
"TrustedProxyAddresses": ["127.0.0.1"],
"OperatorAllowedAddresses": ["127.0.0.1"]
},
"Provisioning": {
"Issuer": "final-factory-rendezvous-smoke",
"Audience": "rendezvous-service",
"ClockSkewSeconds": 30,
"SigningKeys": [
{
"KeyId": "local-smoke-1",
"SecretReference": "file:/run/secrets/rendezvous-signing-key",
"CredentialKinds": ["DedicatedPublisher"],
"GameId": "space-game",
"EnvironmentId": "smoke",
"NotBefore": "2026-01-01T00:00:00Z",
"SignUntil": "2100-01-01T00:00:00Z",
"VerifyUntil": "2100-01-02T00:00:00Z"
}
],
"Games": [
{
"GameId": "space-game",
"EnvironmentId": "smoke",
"Enabled": true,
"ProtocolVersions": [1, 2],
"Regions": ["local"],
"VisibilityModes": ["Public"],
"PublisherTrustModes": ["ManagedDedicated"],
"MetadataValueMaxBytes": {
"mode": 32
},
"RequiredMetadataKeys": [],
"MetadataMaxBytes": 512,
"MetadataMaxKeys": 1,
"MaxListingsPerPrincipal": 10,
"MaxAnonymousListingsPerAddress": 0,
"MaxActiveJoinAttempts": 100,
"FallbackPolicy": "DedicatedEndpointAllowed"
}
]
}
}
}
+35
View File
@@ -0,0 +1,35 @@
name: rendezvous-local
services:
rendezvous:
image: "${RENDEZVOUS_IMAGE:-finalfactory/rendezvous:local}"
build:
context: ../..
dockerfile: Dockerfile
init: true
user: "${RENDEZVOUS_UID:?set RENDEZVOUS_UID to a non-root host UID}:${RENDEZVOUS_GID:?set RENDEZVOUS_GID to its GID}"
read_only: true
tmpfs:
- /tmp:rw,noexec,nosuid,nodev,size=16m,uid=${RENDEZVOUS_UID},gid=${RENDEZVOUS_GID},mode=0700
cap_drop:
- ALL
security_opt:
- no-new-privileges:true
pids_limit: 128
mem_limit: 512m
cpus: 1.0
ulimits:
nofile:
soft: 4096
hard: 4096
stop_grace_period: 40s
restart: unless-stopped
environment:
ASPNETCORE_ENVIRONMENT: Production
ASPNETCORE_HTTP_PORTS: "8080"
volumes:
- ./appsettings.Production.json:/app/appsettings.Production.json:ro
- ./secrets/signing-key:/run/secrets/rendezvous-signing-key:ro
ports:
- "127.0.0.1:8080:8080/tcp"
- "9050:9050/udp"
+2
View File
@@ -0,0 +1,2 @@
*
!.gitignore
+48
View File
@@ -0,0 +1,48 @@
[Unit]
Description=Final Factory Rendezvous service
Documentation=https://git.finalfactory.de/HeiKyu/Rendezvous
After=network-online.target time-sync.target
Wants=network-online.target time-sync.target
[Service]
Type=simple
User=rendezvous
Group=rendezvous
WorkingDirectory=/opt/rendezvous
ExecStart=/usr/bin/dotnet /opt/rendezvous/FinalFactory.Rendezvous.Server.dll
Environment=ASPNETCORE_ENVIRONMENT=Production
Environment=ASPNETCORE_HTTP_PORTS=8080
Environment=DOTNET_EnableDiagnostics=0
EnvironmentFile=-/etc/rendezvous/rendezvous.env
Restart=on-failure
RestartSec=5s
KillSignal=SIGTERM
KillMode=mixed
TimeoutStopSec=40s
NoNewPrivileges=true
PrivateDevices=true
PrivateTmp=true
ProtectClock=true
ProtectControlGroups=true
ProtectHome=true
ProtectHostname=true
ProtectKernelLogs=true
ProtectKernelModules=true
ProtectKernelTunables=true
ProtectSystem=strict
RestrictAddressFamilies=AF_UNIX AF_INET AF_INET6
RestrictNamespaces=true
RestrictRealtime=true
RestrictSUIDSGID=true
CapabilityBoundingSet=
AmbientCapabilities=
LockPersonality=true
SystemCallArchitectures=native
UMask=0077
LimitNOFILE=4096
CPUQuota=200%
MemoryMax=2G
TasksMax=128
[Install]
WantedBy=multi-user.target
File diff suppressed because it is too large Load Diff
@@ -129,9 +129,18 @@ until the owner records:
- which games may enable anonymous unlisted player hosting;
- deployment regions, data-processing jurisdiction, and approval of the stated
30-day audit/13-month aggregate retention periods;
- the per-game dedicated fallback endpoint policy;
- the measured supported profile and whether the 99.5% single-active objective
is sufficient or shared-state/high-availability work must be brought forward.
- the per-game dedicated fallback endpoint policy.
Issue #18 measured and ratified the original 2-vCPU/2-GiB, 25,000-listing,
10,000-attempt core-state candidate profile and retained the 99.5% single-active
topology. It does not claim that core measurements prove public HTTP/UDP SLOs.
The versioned evidence, RTO, failure domains, and explicit signals that trigger
shared-state/high-availability work are recorded in the
[capacity and resilience gate](../operations/capacity-and-resilience.md). The
real-network canary in #23 must confirm that the proposed regional launch load
fits this profile and validate the public SLOs; it may lower the launch cap but
may not silently enable a
second active instance.
These are configuration and launch decisions, not permission to weaken the
tenant, replay, endpoint-verification, or secret-handling controls.
+4 -2
View File
@@ -14,5 +14,7 @@ The public .NET types live in `FinalFactory.Rendezvous.Contracts`, target
vectors and a public-API snapshot make accidental wire or source compatibility
changes fail the normal test gate.
Any incompatible change requires a new contract version. Additive JSON fields
may be introduced within v1 because v1 readers ignore unknown object members.
Readers ignore unknown JSON members, but the release gate deliberately treats
any accepted OpenAPI or golden JSON surface drift as a contract-version change.
That conservative policy makes additive and incompatible published changes
equally visible to consumers instead of relying on an undocumented minor shape.
+233
View File
@@ -0,0 +1,233 @@
# Secure single-active Linux deployment
Tracking: #17
Rendezvous v1 stores listings, observed endpoints, join attempts, replay markers,
and runtime revocations only in the process that accepted them. Deploy exactly
one active instance. A second live replica would have a different directory and
replay boundary; `SingleActiveInstance=false` is therefore rejected rather than
presented as high availability.
## Pinned container
The root `Dockerfile` uses a multi-stage .NET 10 build and pins both Microsoft
base images by multi-architecture manifest digest. The runtime is the chiseled
ASP.NET image, contains only the published server, runs as UID/GID 1654, exposes
TCP 8080 and UDP 9050 explicitly, and does not require a writable application
directory. Supply a small writable `/tmp` tmpfs because runtime libraries can
legitimately need temporary space; keep the root filesystem read-only.
From a clean checkout:
```bash
docker build --pull=false --tag finalfactory/rendezvous:local .
docker inspect --format '{{.Config.User}}' finalfactory/rendezvous:local
```
The reported user must be `1654:1654`. Digest pins make a rebuild reproducible;
updating .NET is an explicit reviewed change to the tag, digest, SDK pin, and
lock files together. Do not replace the digest with `latest` in production.
The local Compose example applies a read-only root, non-root user, no Linux
capabilities, `no-new-privileges`, bounded PIDs/files/memory/CPU, and a shutdown
grace period longer than the service drain deadline:
```bash
install -d -m 0700 deploy/compose/secrets
umask 077
openssl rand -out deploy/compose/secrets/signing-key 32
export RENDEZVOUS_UID="$(id -u)"
export RENDEZVOUS_GID="$(id -g)"
test "$RENDEZVOUS_UID" -ne 0
docker compose -f deploy/compose/compose.yaml up --build --detach
```
`deploy/compose/appsettings.Production.json` is a local/private-bridge smoke
profile, not an Internet template: TCP is published only on host loopback, the
explicit `rendezvous` host name serves isolated clients on the Compose network,
and the profile deliberately opts into private advertised endpoints without a
TLS proxy. Its random key is ignored by Git and must be
deleted after use. Its deliberately long key window only keeps this disposable
local fixture usable; production keys require short, reviewed rotation windows.
Production configuration must use its real public names and must leave
`AllowPrivatePublicEndpoints` false.
## Production topology
Use one active service behind a source-preserving edge:
```text
clients -- HTTPS/443 --> TLS reverse proxy -- HTTP/8080 --> Rendezvous
clients -- UDP/9050 -------------------------------------> Rendezvous
```
- Give the HTTPS origin and UDP endpoint stable DNS names. Set
`PublicHttpBaseUrl` to the exact external HTTPS origin and `PublicUdpHost` /
`PublicUdpPort` to the endpoint given to game clients.
- Terminate TLS 1.2 or newer at a maintained reverse proxy. Bind internal HTTP
only to the private proxy network. Restrict `AllowedHosts` to the public HTTP
host; wildcard host filtering is rejected.
- Put only the proxy's exact literal addresses in
`Rendezvous:AbuseProtection:TrustedProxyAddresses`. Rendezvous ignores
forwarded headers from every other source. Keep the last proxy from replacing
the original client address and prevent direct access to TCP 8080.
- Forward UDP as UDP, without an HTTP proxy. NAT, load balancer, firewall, and
return routing must preserve the client's source IP/port and must send replies
from the same advertised IP/port. Many HTTP load balancers, Kubernetes ingress
controllers, rootless container port proxies, anycast products, and generic
L7 services cannot guarantee this. Do not deploy through one unless an actual
host/join smoke proves both observed source and reply path. A load balancer
must have exactly one healthy Rendezvous target.
- Permit inbound TCP 443 to the TLS proxy and UDP 9050 to Rendezvous. Permit the
proxy to reach TCP 8080. Permit DNS, time synchronization, image/telemetry
destinations as required by local policy, and UDP replies to client endpoints.
Deny public TCP 8080 and every unused inbound port.
Readiness is the load-balancer gate; liveness is only a process-health signal.
Remove a draining instance from new traffic when `/health/ready` becomes 503.
Do not use liveness failure to start a second active process while the old one
still owns the public UDP address.
## Required production configuration
Production startup validates all of these before binding listeners:
- an absolute path-free HTTPS `PublicHttpBaseUrl`;
- an unambiguous public `PublicUdpHost` and port;
- `SingleActiveInstance=true`, an explicit non-wildcard `AllowedHosts`, and at
least one exact trusted TLS-proxy address;
- a 1-30 second drain deadline whose minimum observation interval is shorter;
- at least one enabled game policy and an active scoped signing key.
Missing values produce an actionable startup error. The checked-in base file is
intentionally unsafe for Production so an accidental bare launch fails closed.
Signing keys support two external references:
- `env:NAME` reads 1-4096 bytes encoded as base64 from `NAME`;
- `file:/absolute/path` reads 1-4096 raw bytes from a non-symlink file.
Prefer a read-only container secret owned by the configured container identity.
For systemd, use a root-owned, `rendezvous`-group-owned `0440` file (or an
equivalent narrow ACL) so the non-root process can read but not replace it. A
signing key must contain at least 32 random bytes. Never put the key, publisher/operator
credential, or secret value in JSON, a command argument, an image layer, Compose
environment, logs, metrics, or source control. Configuration contains only the
reference and non-secret lifecycle metadata. A vault/KMS adapter can replace the
provider where local policy requires it.
Keep the host clock synchronized with authenticated NTP. Credential and key
windows use wall time; lease, timeout, drain, and rate-limit deadlines use a
monotonic clock. Alert on clock synchronization loss before rotating keys.
The checked-in Compose limits (one CPU and 512 MiB) are for its isolated smoke
profile, not a production capacity claim. The measured core-state candidate
uses 2 vCPU and 2 GiB with the same 128-PID/4096-descriptor ceilings; see
the [capacity and resilience gate](../operations/capacity-and-resilience.md).
Measure real traffic, then change resource limits and server budgets together.
Memory pressure or CPU throttling must not extend orchestrator termination past
`DrainDeadlineSeconds` plus five seconds.
## Graceful shutdown
SIGTERM and the authenticated operator drain both stop new registrations and
join attempts immediately. On process shutdown, HTTP and UDP remain available
long enough for existing join attempts to finish. The service exits as soon as
the minimum drain interval has elapsed and no attempts remain, or forcibly
clears all ephemeral state at the configured deadline. It then stops UDP and
HTTP listeners and exits. Configure Docker/systemd/Kubernetes termination grace
strictly longer than the service deadline; the examples use 40 seconds for a
30-second drain.
Never use SIGKILL for a normal rollout. After stopping, verify the process is
gone and neither `8080/tcp` nor `9050/udp` is bound before starting its
replacement on the same host. A crashed or force-killed process cannot drain;
clients recover through bounded retries and hosts re-register.
## systemd alternative
Publish the server for Linux, install the immutable output at `/opt/rendezvous`,
place production configuration beside the application read-only, place key
files below `/etc/rendezvous`, and install `deploy/systemd/rendezvous.service`:
```bash
dotnet publish src/FinalFactory.Rendezvous.Server \
--configuration Release --runtime linux-x64 --self-contained false \
--output publish/rendezvous
systemd-analyze verify deploy/systemd/rendezvous.service
sudo systemctl daemon-reload
sudo systemctl enable --now rendezvous.service
```
Create the dedicated `rendezvous` user without a login shell. Keep
`/opt/rendezvous` and `/etc/rendezvous` root-owned and non-writable by that user;
install each required key with `root:rendezvous` ownership and mode `0440`. The
unit applies the measured 2-vCPU/2-GiB core-state candidate profile plus the
same filesystem, privilege, network-family, and shutdown hardening as Compose.
## HTTP and UDP smoke
Build the diagnostic once, then exercise the actual published HTTP and UDP
paths. The test creates a public listing, sends authenticated presence and punch
traffic through UDP 9050, establishes peer-to-peer traffic, reports the outcome,
and deregisters cleanly:
```bash
dotnet build src/FinalFactory.Rendezvous.TestClient --configuration Release
./scripts/smoke-deployment.sh
```
For the local Compose profile, the script derives a ten-minute diagnostic
publisher credential from the ignored local key without printing either secret.
The fixed-scope helper used by the smoke can also support the manual
[TestClient local flow](../integration/test-client.md); it is deliberately not a
production issuer.
For production, do not copy the signing key to the smoke host. Instead inject a
short-lived, region-scoped credential through
`RENDEZVOUS_PUBLISHER_CREDENTIAL`, and set the external endpoints:
```bash
export RENDEZVOUS_PUBLISHER_CREDENTIAL='<short-lived deployment credential>'
export RENDEZVOUS_SMOKE_HTTP_URL='https://rendezvous.your-company.tld/'
export RENDEZVOUS_SMOKE_UDP_ENDPOINT='rendezvous-udp.your-company.tld:9050'
export RENDEZVOUS_SMOKE_GAME_ID='<credential game ID>'
export RENDEZVOUS_SMOKE_ENVIRONMENT_ID='<credential environment ID>'
export RENDEZVOUS_SMOKE_REGION='<credential region>'
export RENDEZVOUS_SMOKE_PROTOCOL_VERSION='<enabled protocol version>'
./scripts/smoke-deployment.sh
```
Those four scope values must match both the short-lived credential and an
enabled server policy. The defaults (`space-game`, `smoke`, `local`, protocol
`1`) are only for the checked-in local Compose profile.
The smoke fails unless both health endpoints and the complete authenticated UDP
mediation/direct-traffic flow succeed. It does not prove every consumer NAT;
run the topology harness and representative external-network tests as well.
## Restart, upgrade, rollback, and backup
Rendezvous has no durable runtime database. Restarting intentionally loses all
listings, observed endpoints, attempts, replay markers, and runtime-only
revocations. Hosts must treat registration as a renewable lease and re-register
after service recovery. Clients must re-browse and start a new bounded attempt.
Back up only reviewed configuration, policy, secret references, key material and
its custody/lifecycle records, deployment manifests, and image digest. Never
claim a backup contains live sessions or endpoints. Restore keys only through the
secret system, not into the image or repository.
For an upgrade:
1. Build and test the new pinned digest; validate configuration without starting
a second active instance.
2. Drain and stop the current process, verify both sockets are released, then
start the replacement on the same public endpoints.
3. Require live/readiness and HTTP+UDP smoke success; monitor host
re-registration, error rate, and direct-connect outcomes.
For rollback, repeat the same stop-before-start sequence with the previously
recorded image digest and compatible configuration/key set. Never run old and
new versions concurrently to avoid split ephemeral state. If a wire-incompatible
change ever becomes necessary, use a new API/protocol version rather than a
rolling two-version replica set.
@@ -0,0 +1,148 @@
{
"schemaVersion": 2,
"evidenceVersion": "v2",
"generatedAt": "2026-07-16T14:10:53.6981858+00:00",
"profile": "candidate",
"runtime": {
"framework": ".NET 10.0.9",
"operatingSystem": "CachyOS",
"kernel": "Unix 7.1.3.2",
"architecture": "X64",
"cpuModel": "AMD Ryzen 7 9800X3D 8-Core Processor",
"processorCount": 2,
"cpuAffinity": "0,1",
"cpuQuota": "not-enforced",
"memoryLimit": "not-enforced",
"garbageCollector": "workstation",
"commitSha": "cf14836d48b0b4aaa67f99433f4fba3585bcd2bb",
"treeState": "clean",
"command": "RENDEZVOUS_CAPACITY_PROFILE=candidate RENDEZVOUS_CAPACITY_CPUSET=0,1 ./scripts/run-capacity-gate.sh",
"imageDigest": "not-containerized",
"workloadSeed": "fixed-sequences-random-identifiers",
"capacityPhaseAverageCpuPercent": 56.37724115383554,
"peakWorkingSetBytes": 169705472,
"managedBytesAfterCleanup": 35615200
},
"targets": {
"visibleListings": 25000,
"activeJoinAttempts": 10000,
"coreControlOperationsPerSecond": 200,
"coreMediationOperationsPerSecond": 2000,
"coreControlP95Milliseconds": 200,
"coreMediationP95Milliseconds": 100,
"maximumAverageCpuPercent": 70,
"maximumWorkingSetBytes": 1610612736,
"soakCycles": 1000,
"soakDurationSeconds": 300
},
"measurements": [
{
"operation": "registration-and-presence",
"samples": 1000,
"p50Milliseconds": 0.003,
"p95Milliseconds": 0.0046,
"p99Milliseconds": 0.0054,
"operationsPerSecond": 282453.96000451926,
"minimumOperationsPerSecond": 200,
"budgetMilliseconds": 200,
"passed": true
},
{
"operation": "lease-renewal",
"samples": 1000,
"p50Milliseconds": 0.0004,
"p95Milliseconds": 0.0009,
"p99Milliseconds": 0.0021,
"operationsPerSecond": 968992.2480620155,
"minimumOperationsPerSecond": 200,
"budgetMilliseconds": 200,
"passed": true
},
{
"operation": "visible-session-browse",
"samples": 250,
"p50Milliseconds": 0.9046,
"p95Milliseconds": 3.3704,
"p99Milliseconds": 3.9471,
"operationsPerSecond": 695.5799787597697,
"minimumOperationsPerSecond": 200,
"budgetMilliseconds": 200,
"passed": true
},
{
"operation": "join-attempt-issuance",
"samples": 1000,
"p50Milliseconds": 0.0029,
"p95Milliseconds": 0.0045,
"p99Milliseconds": 0.0055,
"operationsPerSecond": 296428.042092782,
"minimumOperationsPerSecond": 200,
"budgetMilliseconds": 200,
"passed": true
},
{
"operation": "simultaneous-punch-pairing",
"samples": 1000,
"p50Milliseconds": 0.0043,
"p95Milliseconds": 0.0073,
"p99Milliseconds": 0.0115,
"operationsPerSecond": 109212.03516627532,
"minimumOperationsPerSecond": 2000,
"budgetMilliseconds": 100,
"passed": true
},
{
"operation": "principal-revocation",
"samples": 50,
"p50Milliseconds": 0.518,
"p95Milliseconds": 0.7049,
"p99Milliseconds": 11.8557,
"operationsPerSecond": 1320.1773262184577,
"minimumOperationsPerSecond": 50,
"budgetMilliseconds": 200,
"passed": true
},
{
"operation": "telemetry-recording",
"samples": 1000,
"p50Milliseconds": 0.0001,
"p95Milliseconds": 0.0001,
"p99Milliseconds": 0.0001,
"operationsPerSecond": 1438641.9220256077,
"minimumOperationsPerSecond": 10000,
"budgetMilliseconds": 1,
"passed": true
},
{
"operation": "coincident-listing-attempt-expiry",
"samples": 1,
"p50Milliseconds": 29.6882,
"p95Milliseconds": 29.6882,
"p99Milliseconds": 29.6882,
"operationsPerSecond": 33.682962483916384,
"minimumOperationsPerSecond": 0,
"budgetMilliseconds": 200,
"passed": true
}
],
"state": {
"peakListings": 25000,
"peakAttempts": 10000,
"peakReplayMarkers": 0,
"finalListings": 0,
"finalAttempts": 0,
"finalReplayMarkers": 0,
"expiryChurn": 94906,
"maintenanceSweeps": 36307,
"soakCyclesCompleted": 75126848,
"soakDurationSeconds": 300.0000015,
"soakPeakScheduledExpiryEntries": 7,
"soakManagedGrowthBytes": -257288,
"soakHandleGrowth": 2,
"restartStartedEmpty": true,
"overloadWasTyped": true,
"recoverySucceeded": true
},
"failures": [],
"passed": true
}
+86
View File
@@ -0,0 +1,86 @@
{
"schemaVersion": "1.0",
"recordedAt": "2026-07-16",
"issue": 21,
"consumerIssue": "Kyuubi/SpaceGame#3",
"result": "checkpoint-pass-with-external-gates",
"rendezvousBaseCommit": "ebb5eb617c0bbb170418afab396b68584b7f992e",
"consumerCommit": "f3f5bc29810c362656cd7143bec1ddc2cfaf9f22",
"consumerIssueComment": 11469,
"packages": {
"FinalFactory.Rendezvous.Client": {
"version": "1.0.0",
"source": "local-candidate",
"sourceCommit": "07004cd75fe172aa5dfdb3edda22fc280a4c4477",
"sha256": "fb156cf48b49f75c244dd25ea7cc4aa9fc6fab0a878393bb7efd5d9b131d0395"
},
"FinalFactory.Rendezvous.Contracts": {
"version": "1.0.0",
"source": "local-candidate",
"sourceCommit": "07004cd75fe172aa5dfdb3edda22fc280a4c4477",
"sha256": "a82ba986d3905d599096d1d8ce8f32cd4feb104abfca37b0f65e0d2ef3df9a6f"
},
"LiteNetLib": {
"version": "2.1.4"
}
},
"localRun": {
"processes": ["Rendezvous", "Godot SpaceGame host", "two sequential Godot SpaceGame clients"],
"typedOutcome": "Connected",
"gameAdmission": "Accepted",
"directGameplay": true,
"authenticatedSessions": 2,
"directInputs": 2,
"directSnapshots": 2,
"lifecyclePackets": 4,
"gameplayTransport": "caller-owned-litenetlib",
"rendezvousGameplayPayloadPath": "none",
"hostLeaseRenewed": true,
"reconnected": true,
"deregistered": true
},
"linuxRun": {
"runtime": "Godot 4.7 .NET Linux x86_64",
"freshExport": true,
"sourceDirty": false,
"optimized": true,
"dedicatedHostNamespace": "docker",
"remoteClientNamespace": "docker",
"topology": "private-bridge",
"directGameplay": true,
"fallback": "PunchTimedOut to explicit Docker-gateway endpoint, then Accepted game admission and direct gameplay",
"artifactHashes": "SpaceGame issue #3 comment 11469"
},
"negativePaths": {
"incompatibleProtocol": "proven",
"staleHostPresence": "proven",
"punchTimeout": "proven",
"invalidAdmission": "integration-proven",
"capacity": "regression-tested",
"fallbackConnection": "godot-and-isolated-linux-proven",
"reconnect": "proven"
},
"verification": {
"debugBuild": "passed",
"releaseBuild": "passed",
"debugTests": { "passed": 31, "failed": 0 },
"releaseTests": { "passed": 31, "failed": 0 },
"exportRelease": "optimized-without-debug-symbols",
"format": "passed",
"shellcheck": "passed",
"godotReconnectHarness": "passed",
"godotFallbackHarness": "passed",
"linuxContainerHarness": "passed-clean-source",
"failureMatrix": "passed",
"adversarialReview": "passed-after-fixes"
},
"openGates": [
"public-package-restore",
"representative-external-nat"
],
"relatedSpaceGameGates": [
"production-enet-replacement",
"capacity-profiles-64-and-128",
"sigterm-drain-save"
]
}
+232
View File
@@ -0,0 +1,232 @@
# Game integration seams
Tracking: #20
Use the [TestClient start-to-finish guide](test-client.md) before integrating a
game. It proves the service and network path without engine or game code. This
page documents only the seams that the diagnostic cannot choose for a game:
package/version policy, ownership of the gameplay socket, host admission,
metadata, credential custody, and deployment compatibility.
## Packages and compatibility
Consume `FinalFactory.Rendezvous.Client` and
`FinalFactory.Rendezvous.Contracts` from the approved Gitea NuGet source and pin
both to the same exact released version. Do not use a floating version range.
The current release matrix is machine-readable in
[`compatibility.json`](../releases/compatibility.json); the same window is
available from authenticated `GET /v1/operator/status`.
The authoritative package feed is
`https://git.finalfactory.de/api/packages/HeiKyu/nuget/index.json`. Add it to the
consumer's `NuGet.config` and map only Rendezvous packages to it; retain the
consumer's existing NuGet.org mapping for other dependencies:
```xml
<packageSources>
<add key="FinalFactory" value="https://git.finalfactory.de/api/packages/HeiKyu/nuget/index.json" />
</packageSources>
<packageSourceMapping>
<packageSource key="FinalFactory">
<package pattern="FinalFactory.Rendezvous.*" />
</packageSource>
<packageSource key="nuget.org">
<package pattern="*" />
</packageSource>
</packageSourceMapping>
```
When the feed is anonymously readable, no reader credential is needed. If
registry policy requires authentication, use the platform's NuGet credential
provider or a protected per-user/CI NuGet configuration populated by the secret
manager. Never put a registry token in the project file, repository,
package-source URL, or `dotnet` command argument.
```xml
<ItemGroup>
<PackageReference Include="FinalFactory.Rendezvous.Client" Version="1.0.0" />
<PackageReference Include="FinalFactory.Rendezvous.Contracts" Version="1.0.0" />
</ItemGroup>
```
Run `dotnet restore`, then `dotnet list package --include-transitive` and verify
that Client and Contracts resolve to the same exact version and LiteNetLib to the
release matrix version before compiling the game.
Release 1.0.0 targets `netstandard2.1`, requires LiteNetLib `2.1.4`, speaks HTTP,
UDP, and connection-ticket contract version `1`, and requires an exact
tenant-configured gameplay protocol match. A package patch does not silently
change a wire version. Follow the [release and migration policy](../releases/README.md)
when changing any dimension, and validate the generated
[OpenAPI v1 document](../api/rendezvous-v1.json) rather than hand-building HTTP.
## One caller-owned gameplay socket
Create the game's LiteNetLib manager through `RendezvousNetListener`; do not open
a separate NAT socket. The game owns start, stop, and disposal. A coordinator
owns polling while it is active, so call its `Poll()` once from the game/network
thread and do not also call `NetManager.PollEvents()` during that period.
```csharp
RendezvousNetListener networkEvents = new();
NetManager gameplayNetwork = networkEvents.CreateManager();
gameplayNetwork.ChannelsCount = 3; // set the game's required count before Start
if (!gameplayNetwork.Start(gameplayPort))
{
throw new InvalidOperationException("Gameplay UDP socket could not start.");
}
using RendezvousHostCoordinator host = new(
gameplayNetwork,
networkEvents,
mediatorEndPoint,
publishedSession,
joinClient);
host.Poll(); // call each game frame while this coordinator owns polling
```
LiteNetLib defaults to one QoS channel. Set `ChannelsCount` before `Start` when
the game protocol uses additional channels; both peers must configure the same
count. Rendezvous does not choose, remap, or reserve a gameplay channel.
Register normal game callbacks on `networkEvents.GameplayEvents`. Rendezvous
reserves only its authenticated direct requests and forwards other callbacks.
The same socket sends host presence, punches through the mediator, establishes
the peer, and then carries gameplay. A NAT introduction is not success; accept a
peer only after the coordinator reports the typed `Connected` outcome.
`Poll()` does not fetch new invitations. Schedule
`RefreshJoinAttemptsAsync` repeatedly for the entire hosting lifetime using a
bounded caller-owned timer (the diagnostic uses 250 ms), never allow two refreshes
to overlap, and inspect each typed result. The refresh performs HTTP work and
queues a snapshot; it does not call the LiteNetLib manager. Continue calling
`Poll()` on the manager's owning thread so the queued snapshot, presence traffic,
and callbacks are processed. Run the lease maintainer concurrently and cancel
both loops before disposing the coordinator.
For example, start one sequential refresh loop when hosting begins and await it
during shutdown:
```csharp
static async Task RefreshInvitationsAsync(
RendezvousHostCoordinator host,
CancellationToken cancellationToken)
{
using PeriodicTimer timer = new(TimeSpan.FromMilliseconds(250));
do
{
RendezvousClientResult<int> result =
await host.RefreshJoinAttemptsAsync(cancellationToken);
if (!result.IsSuccess)
{
ObserveBoundedHostRefreshFailure(result.Error);
}
}
while (await timer.WaitForNextTickAsync(cancellationToken));
}
```
On the joining side, create an attempt through `RendezvousJoinClient`, then give
the issued attempt to `RendezvousClientCoordinator` using the same manager and
listener. Cancellation, outcome reporting, bounded deadlines, fallback, and
lease-maintainer examples are in the packaged
[`FinalFactory.Rendezvous.Client` README](../../src/FinalFactory.Rendezvous.Client/README.md).
## Host admission remains game-owned
The coordinator privately validates and consumes the signed one-time connection
ticket before accepting the LiteNetLib transport request. Do not create a second
`ConnectionTicketValidator` beside it: the coordinator deliberately does not
expose the expected or presented ticket. A connected transport proves only that
Rendezvous authorized one attempt; it does not prove player identity,
entitlement, capacity, ban status, or gameplay compatibility.
Treat `AttemptCompleted` with a successful outcome and non-null `Peer` as the
start of game-owned admission. Keep that peer outside authoritative gameplay
until the game's normal authentication and admission exchange succeeds; disconnect
it on rejection or timeout:
```csharp
host.AttemptCompleted += (_, completed) =>
{
if (!completed.Outcome.IsSuccess || completed.Peer is null)
{
return;
}
BeginBoundedGameAuthentication(
completed.Peer,
onAccepted: AdmitToAuthoritativeGameplay,
onRejected: peer => peer.Disconnect());
};
```
Revoke an attempt when the game cancels it. Never log a ticket or capability. A
successful Rendezvous check must not bypass the game's authentication or
authoritative server rules. `ConnectionTicketValidator` is a lower-level
primitive for a custom transport integration that owns the complete request
acceptance path; it is not an extra gate for `RendezvousHostCoordinator`.
## Provision each game and environment
Provision game/environment scope before issuing credentials. The policy fixes
enabled regions, exact gameplay protocols, visibility and publisher trust modes,
metadata schema and byte budgets, quotas, and whether a dedicated fallback may
be published. Unknown or disabled scope fails closed. Follow
[game provisioning and signing-key lifecycle](../security/provisioning.md) for
the complete schema, principal kinds, secret providers, overlap, and revocation.
Dedicated publisher credentials belong only on trusted hosting infrastructure.
Never ship one in a player build, repository, image layer, appsettings file, URL,
argument, log, crash report, or analytics event. Issue a short-lived credential
scoped to one game/environment and its allowed regions from the trusted
deployment boundary. Player-host grants are issued to an authenticated player
session at runtime and are never embedded in the build. Player-host grants,
dedicated publishers, anonymous unlisted hosts, and operators are separate
principal kinds; do not interchange them.
Rotate signing keys with an overlap:
1. install a new authorized key inside its `NotBefore`/`SignUntil` window;
2. begin issuing with it while the old key remains verify-only;
3. wait at least the maximum credential lifetime plus allowed clock skew;
4. retire the old verifier after `VerifyUntil` and preserve custody records.
A suspected compromise is not routine rotation: stop issuance, revoke the exact
key through the protected operator route, remove or replace it in provisioning,
invalidate affected credentials, and follow the
[key-compromise runbook](../operations/incident-runbooks.md#signing-key-or-issuer-compromise).
## Metadata is public and policy-owned
Treat listing metadata as untrusted public input. Define a small allowlist in
each provisioned game's `MetadataValueMaxBytes`, set `RequiredMetadataKeys`, and
keep `MetadataMaxKeys` and `MetadataMaxBytes` to the smallest useful values. Values
must be display data only—for example a bounded map or ruleset identifier. Never
publish player identity, free-form chat, secrets, access tokens, internal
addresses, world state, or data needed for authoritative gameplay.
The platform contract caps metadata at 32 keys, 256 UTF-8 bytes per value, and
4096 encoded bytes total; tenant policy can and should be smaller. Build version
and display name are separately bounded public fields. Games must escape metadata
for their UI and must not infer trust from a listing being present.
## Local, staging, and production path
Use the checked-in Compose profile only for the local TestClient guide. For a
real environment:
1. provision the game/environment policy and externally held signing keys;
2. deploy one active service behind the source-preserving HTTPS/UDP topology in
[secure single-active Linux deployment](../deployment/linux.md);
3. install matching exact package versions in the game and set its service and
mediator endpoints through environment-specific configuration;
4. pass the TestClient health/publish/browse/punch/direct-traffic smoke using a
short-lived diagnostic credential;
5. run the topology harness and representative consumer-network trials; and
6. monitor typed outcomes and bounded metrics before broad rollout.
Rendezvous v1 has no relay, account system, matchmaking engine, server-browser
UI, gameplay authority, or durable session database. A game owns player-facing
recovery and an explicit fallback. Do not describe direct traversal as guaranteed.
+112
View File
@@ -0,0 +1,112 @@
# SpaceGame consumer pilot
Tracking: Rendezvous #21 and SpaceGame #3.
The current SpaceGame checkpoint proves that the v1 client boundary establishes
authenticated direct LiteNetLib traffic without taking ownership of the game's
protocol, admission, player identity, entity identity, capacity, lifecycle, or
gameplay payloads. Real Godot processes, reconnect, an explicit dedicated
fallback, and a fresh Linux export now pass. The public package restore and a
representative external NAT/CGNAT canary remain required before #21 can close.
## Pinned checkpoint
| Input | Value |
| --- | --- |
| Rendezvous compatibility source | `ebb5eb617c0bbb170418afab396b68584b7f992e` plus the current #21 configuration/evidence changes |
| Rendezvous package source | `07004cd75fe172aa5dfdb3edda22fc280a4c4477` |
| SpaceGame source | `f3f5bc29810c362656cd7143bec1ddc2cfaf9f22` |
| Client package | `FinalFactory.Rendezvous.Client` `1.0.0` |
| Contracts package | `FinalFactory.Rendezvous.Contracts` `1.0.0` |
| LiteNetLib | `2.1.4` |
| HTTP, UDP, ticket contracts | `1` |
| SpaceGame gameplay protocol | `2` |
At the checkpoint date, the Final Factory Gitea NuGet service was reachable but
both `FinalFactory.Rendezvous.*` `1.0.0` registrations returned HTTP 404. The run
therefore restored locally built candidate packages with the hashes recorded in
[`spacegame.json`](../evidence/consumers/spacegame.json). This proves candidate
compatibility, not immutable registry publication. The release package restore
must be repeated from the public feed.
## Proven local path
The SpaceGame host and client each create one caller-owned `NetManager`, set its
three gameplay QoS channels before `Start`, and give the same manager and
`RendezvousNetListener` to the coordinator. Rendezvous authenticates discovery,
join authorization, host presence, mediation, and connection outcome reporting.
After traversal, SpaceGame performs a separate audience-bound admission exchange
on its own reliable command channel. A trusted game-auth boundary mints the
opaque assertion; the player process never receives the signing key.
The authoritative host rejects expired, replayed, incorrectly signed,
wrong-listing, duplicate-player, over-capacity, identity-mismatched,
out-of-sequence, and over-rate traffic. It assigns a canonical game entity ID
only after admission. The player ID, entity ID, listing ID, join-attempt ID, and
LiteNetLib peer ID remain distinct values.
The bounded real-process harnesses observed:
- host publication and lease maintenance;
- browser compatibility filtering and join authorization;
- typed traversal outcome `Connected`;
- successful audience-bound game admission;
- reliable ordered frame-definition and spawn lifecycle records, reliable
ordered input, and sequenced state snapshots on the caller-owned gameplay
socket;
- disconnect and a new authenticated session for the same durable player while
LiteNetLib peers and canonical entity IDs change;
- immediate host lease renewal and successful host deregistration;
- a fresh optimized Linux export running the host and client in distinct
hardened container namespaces; and
- a forced punch timeout that connects the isolated client to an explicitly
advertised, non-loopback Docker-gateway fallback and repeats game admission.
Rendezvous exposes no gameplay relay API; all lifecycle, command, and snapshot
bytes are sent by SpaceGame through its caller-owned `NetManager`. Both Debug
and Release builds passed. Both Debug and Release test runs passed 31 tests with
zero failures. ExportRelease is optimized with debug symbols removed. The
focused formatter, shell checker, fresh-export provenance gate, clean
candidate-package restore, and adversarial branch review also passed.
## Failure evidence
| Path | Evidence | Status |
| --- | --- | --- |
| Incompatible protocol | protocol `999` returns no compatible listing and starts no traversal | Proven |
| Stale/no host presence | typed `NoHostPresence/RendezvousService/HostPresence/Mediation` | Proven |
| Traversal timeout | non-listening mediator produces typed `PunchTimedOut/LocalTraversal/NatTraversal/NatTraversal` | Proven |
| Rejected game admission | invalid signature denies gameplay in the process matrix; wrong audience, expiry, and replay are regression-tested | Proven |
| Capacity and duplicate player | game-owned roster rejects both and publishes current capacity | Regression-tested |
| Configured fallback | typed `PunchTimedOut`, explicit non-loopback endpoint, same game admission, direct gameplay | Proven locally and across Linux namespaces |
| Disconnect | host observes zero active players and final admitted count zero | Proven |
| Reconnect | same durable player enters a second authenticated session with new peer/entity IDs | Proven |
## Rendezvous-side compatibility fixes
The pilot found generic integration gaps and keeps their fixes in this
repository:
- the local production-shaped smoke tenant accepts SpaceGame gameplay protocol
`2` and the bounded `mode` metadata key;
- the Compose smoke tenant explicitly allows its private-network service name
and enables only the dedicated-endpoint fallback policy;
- SDK guidance requires games using multiple LiteNetLib QoS channels to set
`ChannelsCount` before `Start` and states that Rendezvous reserves no gameplay
channel; and
- the local credential helper rejects any signing-key file with group or other
permissions, in addition to its ownership, symlink, and hard-link checks.
Documentation contract tests cover these generic requirements.
## Remaining acceptance gates
Do not mark #21 passed until both remaining external gates have direct evidence:
1. restore the exact immutable `1.0.0` packages from the public Gitea feed; and
2. run representative external NAT/CGNAT canaries and record the network
topology and typed outcome.
SpaceGame #3 remains open independently for the production ENet replacement,
64/128-player profiles, and SIGTERM/drain/save evidence. The consumer pilot
does not claim those broader game-migration gates.
+222
View File
@@ -0,0 +1,222 @@
# Start-to-finish TestClient guide
Tracking: #20, #25
`FinalFactory.Rendezvous.TestClient` is the supported executable proof that a
consumer can publish, browse, authorize, punch, connect, exchange direct traffic,
and diagnose a failure using only the public Client and Contracts packages. It is
intentionally thin: a polished server browser and player-facing connection UI
belong in each game repository.
> **Traversal boundary:** Rendezvous v1 is not a relay and cannot guarantee a
> connection through symmetric NAT, carrier-grade NAT, restrictive firewalls,
> VPNs, or platform policy. It provides no accounts, social system, skill-based
> matchmaking, gameplay server, gameplay authority, or gameplay transport.
The automated scenario matrix, privileged Linux namespace run, and simulation
limits are in the [deterministic topology harness](topology-harness.md). The
[SDK seam guide](sdk-seams.md) covers the few integration details that this
executable cannot show.
## First local connection from a clean checkout
Prerequisites are the pinned .NET SDK, Docker with Compose, OpenSSL, Python 3,
`curl`, and `jq`. Run these commands from the repository root. The generated key
and credential are disposable local fixtures, not production provisioning.
```bash
install -d -m 0700 deploy/compose/secrets
umask 077
openssl rand -out deploy/compose/secrets/signing-key 32
export RENDEZVOUS_UID="$(id -u)"
export RENDEZVOUS_GID="$(id -g)"
test "$RENDEZVOUS_UID" -ne 0
docker compose -f deploy/compose/compose.yaml up --build --detach
ready=false
for attempt in {1..45}; do
if curl --fail --silent http://127.0.0.1:8080/health/ready >/dev/null; then
ready=true
break
fi
sleep 1
done
test "$ready" = true
curl --fail http://127.0.0.1:8080/health/live
curl --fail http://127.0.0.1:8080/health/ready
dotnet build src/FinalFactory.Rendezvous.TestClient --configuration Release
```
Only the host terminal needs a publisher credential. Disable shell tracing before
capturing it; the helper prints the credential on stdout so command substitution
can place it directly in the environment without writing it to disk.
```bash
set +x
export RENDEZVOUS_PUBLISHER_CREDENTIAL="$(./scripts/mint-local-publisher-credential.sh)"
```
The helper accepts no arguments, reads the ignored `0600` local Compose key, and
mints only `space-game` / `smoke` / `local` / protocol `1` for ten minutes. It is
not a reusable issuer or an example for production. Never put the result in a
command argument, URL, shell history, log, screenshot, support ticket, captured
fixture, or source file.
In terminal 1, publish a host. It stays alive for at most 60 seconds and exits
after a joining peer completes the authenticated echo exchange:
```bash
dotnet run --project src/FinalFactory.Rendezvous.TestClient \
--configuration Release --no-build -- \
host --service http://127.0.0.1:8080/ --mediator 127.0.0.1:9050 \
--game space-game --environment smoke --region local --protocol 1 \
--display-name "Local diagnostic" --timeout-seconds 60 --run-seconds 60 \
--exit-after-echo
```
Copy the public listing ID printed by the host, or discover it from terminal 2:
```bash
dotnet run --project src/FinalFactory.Rendezvous.TestClient \
--configuration Release --no-build -- \
browse --service http://127.0.0.1:8080/ \
--game space-game --environment smoke --region local --protocol 1
```
In terminal 3, either omit `--listing` and select interactively, or provide the
copied ID for deterministic selection:
```bash
dotnet run --project src/FinalFactory.Rendezvous.TestClient \
--configuration Release --no-build -- \
join --service http://127.0.0.1:8080/ --mediator 127.0.0.1:9050 \
--game space-game --environment smoke --region local --protocol 1 \
--listing REPLACE_WITH_LISTING_UUID --timeout-seconds 30
```
Success means the joiner prints `join.connected` and verified direct traffic,
and the host prints verified direct traffic before deregistering. The host and
joiner each create one caller-owned LiteNetLib manager. The same UDP socket sends
presence and punch traffic, accepts the authenticated peer, and carries the
ping/echo/ack/completion payload; direct traffic does not pass through the HTTP
service or mediator.
Clean up secrets and the disposable service when finished:
```bash
unset RENDEZVOUS_PUBLISHER_CREDENTIAL
docker compose -f deploy/compose/compose.yaml down
rm deploy/compose/secrets/signing-key
```
## Script and JSON automation
`--script` forbids prompts and selects the first compatible listing unless
`--listing UUID` fixes the choice. `--json` emits one JSON object per line with
`version: 1`. New optional properties may be added, but event names and exit
codes are stable automation contracts. Informational events use stdout and
failures use stderr.
The deployment smoke performs the full health, publish, join, mediation, direct
traffic, outcome-report, and cleanup flow using bounded waits:
```bash
dotnet build src/FinalFactory.Rendezvous.TestClient --configuration Release
./scripts/smoke-deployment.sh
```
For custom automation, capture JSON and preserve the process status separately:
```bash
set +e
dotnet run --project src/FinalFactory.Rendezvous.TestClient \
--configuration Release --no-build -- \
browse --service http://127.0.0.1:8080/ \
--game space-game --environment smoke --region local --protocol 1 \
--script --json >browse.jsonl
status=$?
set -e
jq -e 'select(.version == 1 and .event == "browse.completed")' browse.jsonl
test "$status" -eq 0
```
Never use an unbounded sleep to orchestrate processes. Wait for versioned events
such as `host.ready` and apply a deadline. Useful success events are
`host.registered`, `host.ready`, `host.direct-traffic`, `host.deregistered`,
`browse.completed`, `browse.session`, `join.connected`, `join.direct-traffic`,
and `join.outcome-report`.
| Exit | Meaning |
| ---: | --- |
| `0` | Requested diagnostic flow completed successfully |
| `2` | Invalid command or options |
| `3` | Missing or invalid local configuration |
| `10` | HTTP, registration, browser, lease, or socket failure |
| `11` | No compatible session was available or selected |
| `12` | Authorization or traversal reached a typed terminal failure |
| `13` | Direct connection succeeded but the direct traffic proof failed |
| `130` | Caller cancellation or Ctrl+C |
## Observe a safe failure
Run this after the protocol-1 browse in terminal 2 and before the terminal-3
join (or restart terminal 1 first). The preceding browse proves that one
protocol-1 host is present. Now browse for deliberately incompatible protocol
`999`. The command emits a successful directory response with
`browse.completed`, `count: 0`, then exits `11` to distinguish compatibility
from a service outage:
```bash
set +e
dotnet run --project src/FinalFactory.Rendezvous.TestClient \
--configuration Release --no-build -- \
browse --service http://127.0.0.1:8080/ \
--game space-game --environment smoke --region local --protocol 999 \
--script --json >incompatible.jsonl
status=$?
set -e
jq -e 'select(.event == "browse.completed" and .phase == "directory" and .count == 0)' \
incompatible.jsonl
test "$status" -eq 11
```
This is a diagnostic failure drill, not a bypass: unknown tenant scope and
protocols still fail closed, and the local helper cannot mint a credential for
them.
## Diagnose by phase, not by guesswork
Start with the exit code, then the last versioned event and its `phase`, `status`,
and typed `outcome`. Endpoint categories may be
reported as `loopback`, `private`, or `public`; raw endpoints, credentials,
capabilities, metadata, and player identities are never emitted.
| Symptom or last event | Distinction | Check next |
| --- | --- | --- |
| `host.configuration`, exit `3` | Local credential variable is missing or malformed before any request | Confirm the named environment variable exists, tracing is off, and the credential has not expired |
| `host.registration`, exit `10` | Publisher authentication, tenant policy, metadata, quota, or HTTP failure | Use the typed status; compare credential scope with game/environment/region and the provisioned policy, then correlate protected server telemetry by operation and time |
| `browse.sessions`, exit `10` | Directory request failed | Check HTTP reachability, `/health/ready`, rate limiting, and contract compatibility |
| `browse.completed` count `0`, or `join.selection` empty, exit `11` | Healthy directory but no compatible visible listing | Match game, environment, region, and exact gameplay protocol; then confirm a host lease is still active |
| Exact `join.selection` failure, exit `10` | Listing disappeared, is hidden, or scope no longer matches | Browse again; do not retry an old listing ID forever |
| `join.authorization`, exit `12` | Service rejected the attempt before NAT traversal | Inspect typed category/outcome for policy, capacity, stale host, or active-attempt limits |
| `join.punch` / `join.traversal`, exit `12` | Mediation or NAT traversal did not establish a peer | Confirm UDP endpoint/reply path, host presence, clocks, firewall/NAT behavior, and topology; use a game-owned fallback if policy supplies one |
| `join.direct-connect`, exit `12` | Introduction occurred but authenticated direct admission failed | Confirm host is polling the same socket, the one-time ticket is current, and game admission did not reject capacity, identity, or bans |
| `join.connected` followed by exit `13` | Peer connected but the direct gameplay-like echo did not finish | Inspect the peer lifecycle and caller polling; this is not an HTTP/directory failure |
Stopping a host without deregistration may leave its listing visible only until
the bounded lease expires. During that window, a join can produce a typed stale
host or traversal outcome; it must not be interpreted as a healthy host. Restarting
the single-active service intentionally loses all ephemeral listings and attempts,
so hosts re-register and clients browse again.
If a terminal outcome reports an authoritative dedicated fallback,
`join.fallback` exposes only availability and endpoint type. TestClient never
connects to it automatically. The game owns the decision, authentication, and
connection policy. If no fallback is present, Rendezvous v1 offers no relay.
## Production use
Do not copy a production signing key to a diagnostic host. Supply a short-lived,
least-scope publisher credential from the deployment secret boundary and set the
external service, mediator, and matching scope variables described in the
[secure Linux deployment smoke](../deployment/linux.md#http-and-udp-smoke).
Run representative external-network tests; loopback success is not NAT coverage.
+91
View File
@@ -0,0 +1,91 @@
# Deterministic topology harness
Issue #14 is verified at three layers. The layers are deliberately separate so
the always-on gate remains deterministic while privileged CI workers can add a
stronger operating-system topology without overstating what local emulation
proves about the public Internet.
## Always-on public-process gate
`TestClientProcessIntegrationTests` launches the built server and the same
`FinalFactory.Rendezvous.TestClient` executable shipped to operators. Every
child process uses `--script --json`, dynamic HTTP and UDP ports, bounded
state-driven waits, and enforced process-tree cleanup.
The suite proves:
| Scenario | Required observation |
| --- | --- |
| Three-party happy path | register, presence-ready, browse, authorize, punch, authenticated LiteNetLib connection, direct ping/echo/ack/completion traffic, outcome report, disconnect, deregister |
| Same-LAN candidate | the connected peer is reported as `loopback` or `private`, never inferred merely from an introduction callback |
| Empty and missing selection | browse exits `11`; exact missing lookup exits `10` |
| Wrong tenant/protocol | no listing is returned for an incompatible protocol; exact joins with either mismatch fail before `join.punch` |
| Traversal timeout | an unreachable mediator produces typed `PunchTimedOut`, exits `12`, advertises the configured dedicated fallback, and never connects to it |
| Caller cancellation | POSIX `SIGINT` exits `130`, deregisters the listing, and removes it from public lookup |
| Abrupt host loss | the listing disappears after the presence window and before its lease expires; public exact lookup intentionally reports `NotFound` |
| Bounded host without a peer | exits `13` and still deregisters |
Captured output is parsed as the stable JSON v1 event schema. Publisher
credentials and signing-key material are checked against all captured output.
The direct traffic payload is handled only by the caller-owned host and client
LiteNetLib managers; the HTTP service and mediator do not implement or observe
the echo protocol.
Run the always-on scenarios with:
```bash
dotnet test Rendezvous.slnx --configuration Release --no-build \
--filter FullyQualifiedName~TestClientProcessIntegrationTests
```
## Deterministic protocol and adverse-state gate
The following real service-boundary tests cover conditions that a public CLI
cannot safely manufacture by accepting raw capabilities or tickets:
| Scenario | Test evidence |
| --- | --- |
| Same-NAT private candidates | `NatMediationProcessorTests.MatchedPeersReceiveOneIntroductionAndSameNatPrivateCandidates` |
| Separate observed endpoints | `NatMediationProcessorTests.DifferentNatsAndInvalidLocalClaimsExposeOnlyObservedPublicEndpoints` |
| One-time introduction and replay | `InMemoryEphemeralRendezvousStoreTests.AttemptCapabilitiesAndIntroductionAreOneTime` |
| Direct ticket replay | `RendezvousCoordinatorIntegrationTests.CallerOwnedManagersCompleteAuthenticatedDirectConnectionAndRejectTicketReplay` |
| Wrong tenant/protocol and stale presence | `InMemoryEphemeralRendezvousStoreTests.JoinRequiresExactScopeProtocolAndFreshHostPresence` |
| Cancellation and late callbacks | `RendezvousCoordinatorBehaviorTests.CancellationCompletesExactlyOnceAndLateCallbacksCannotReopenTheAttempt` |
| Mediator restart | both cases of `UdpMediatorServiceTests.NativeLiteNetLibRequestsIntroduceTheAuthorizedPair`; the restarted case rebinds the same UDP port and completes a native LiteNetLib introduction |
These tests use fake monotonic clocks or state predicates where expiry and race
ordering matter. They do not use fixed sleeps as proof of state.
## Privileged Linux namespace gate
When a Linux CI worker can create network namespaces, the workflow sets
`RENDEZVOUS_RUN_NETNS_TESTS=1` and reruns
`PrivilegedLinuxNatNamespacesCompleteDirectTrafficAcrossSeparateObservedEndpoints`.
The test creates a temporary WAN bridge, an isolated service namespace, two NAT
router namespaces, and isolated host/client LAN namespaces. Each NAT has its own
inside subnet and WAN address. Linux forwarding plus per-router MASQUERADE rules
force the service to observe separate translated endpoints; the public TestClient
processes must then complete authenticated direct traffic through those mappings
using the public candidate. Namespaces, rules, veth pairs, bridge, processes, and
sockets are removed in bounded async-disposal paths. A cleanup failure fails the
test.
If `ip netns add`/`iptables` is unavailable or the worker lacks `CAP_NET_ADMIN`,
CI records the limitation and keeps the always-on loopback suite as the required gate.
To request the privileged run explicitly:
```bash
RENDEZVOUS_RUN_NETNS_TESTS=1 dotnet test Rendezvous.slnx \
--configuration Release --no-build \
--filter FullyQualifiedName~PrivilegedLinuxNatNamespacesCompleteDirectTrafficAcrossSeparateObservedEndpoints
```
## What this does not prove
Loopback, MASQUERADE, and namespace routing cannot reproduce every consumer router,
carrier-grade NAT, firewall, IPv6 transition mechanism, symmetric NAT mapping,
or real-world packet-loss pattern. The separate-observed-endpoint processor
test proves that untrusted private claims are excluded and public candidates are
selected; it is not presented as universal Internet traversal proof. Real
network canaries and measured production readiness remain the scope of issue
#23.
+189
View File
@@ -0,0 +1,189 @@
# Capacity, resilience, and availability gate
Tracking: #18
This gate turns the v1 budgets in ADR 0003 into a repeatable release decision.
It does not turn Rendezvous into a horizontally scalable service: v1 remains one
active process with bounded in-memory state. A second process may be a cold
standby, but it must not accept traffic until the first process has stopped and
released the public HTTP and UDP endpoints.
## Launch envelope and approved core-state profile
The approved core-state profile is one Linux process limited to 2 vCPU and
2 GiB RAM. Public HTTP/UDP numbers are launch objectives that require the #23
real-network canary before they become a supported service claim:
| Dimension | Value | Evidence status |
| --- | --- | --- |
| Visible listings | 25,000 | Enforced and measured here |
| Active join attempts | 10,000 | Enforced and measured here |
| Core control path | 200 operations/second; p95 at most 200 ms | Measured here |
| Core mediation path | 2,000 pairings/second; p95 at most 100 ms | Measured here |
| Sustained HTTP demand | 200 requests/second | #23 launch objective; not yet a supported claim |
| Sustained UDP demand | 2,000 datagrams/second | #23 launch objective; not yet a supported claim |
| Public HTTP/UDP latency | p95 at most 200 ms / 100 ms | #23 launch objective; not yet a supported claim |
| Capacity-phase average CPU / peak working memory | below 70% / below 1.5 GiB | Measured for the core candidate |
| Valid in-profile monthly availability | 99.5%, excluding announced maintenance | Operational objective |
| Process-ready RTO / host-visible recovery | 15 seconds / 90 seconds | 15 seconds automated; 90-second deployment drill required |
The proposed public-network mix is 20% registration/update, 30% lease-critical
renew/delete, 30% browse, and 20% join authorization for HTTP. The UDP mix is
60% authenticated host-presence refresh, 30% attempt contributions, and 10%
invalid or duplicate traffic that must be dropped early. A deployment may use a
lower per-game profile, but must not claim a higher one without new versioned
evidence.
The capacity harness fills the complete state ceilings, then measures
registration plus presence, renewal, a 100-item compatible browse, join
issuance, simultaneous two-peer pairing, principal revocation, and telemetry.
It applies 200/100 ms guardrails and minimum 200 control / 2,000 mediation
operations per second to the core hot path. Those measurements deliberately
exclude Kestrel, LiteNetLib, TLS, JSON, socket scheduling, and the documented
mixed traffic shape. The #23 real-network canary must exercise those layers,
rate-shape the mix, record errors and shedding, and meet the public objectives
before launch; a core result is not a public-network latency or throughput claim.
## Reproduce the evidence
Every push runs the quick profile and the selected fault matrix:
```bash
./scripts/run-capacity-gate.sh
```
Run the production candidate on an otherwise idle Linux host and restrict the
runtime to two logical CPUs. The default candidate includes a five-minute,
high-intensity expiry soak; use 3,600 seconds for a release-candidate endurance
run:
```bash
export RENDEZVOUS_CAPACITY_PROFILE=candidate
export RENDEZVOUS_CAPACITY_CPUSET=0,1
export RENDEZVOUS_CAPACITY_OUTPUT="$PWD/artifacts/capacity/candidate.json"
./scripts/run-capacity-gate.sh
# Release-candidate endurance override:
dotnet run --project tests/FinalFactory.Rendezvous.Capacity \
--configuration Release --no-build -- \
--profile candidate --soak-seconds 3600 \
--output artifacts/capacity/candidate-endurance.json
```
The machine must have at least 2 GiB available to the process. For formal
deployment evidence, run inside the same cgroup/container shape as production.
The v2 JSON embeds the commit and tree state, command, image context, CPU model,
kernel, affinity, cgroup quota/limit, collector mode, and workload seed. Supply
`RENDEZVOUS_EVIDENCE_IMAGE_DIGEST` when running a release image. Do not compare
results collected under a debugger,
concurrent build, thermal throttling, or oversubscribed CI host.
The checked-in baseline is
[`candidate-2cpu.json`](../evidence/capacity/v2/candidate-2cpu.json). It was
produced on .NET 10.0.9/Linux x64 with CPU affinity restricted to two logical
CPUs. It filled 25,000 listings and 10,000 attempts, peaked at about 162 MiB,
and cleared all active/retained state. The five-minute baseline supersedes any
earlier local probe when its timestamp and target duration differ.
## Soak and bounded-state interpretation
Each soak cycle creates a listing, repeatedly renews its lease and refreshes
presence, creates a join attempt, replay marker, and retained outcome, checks
that scheduled expiry entries remain proportional to live keys, then advances
the injected monotonic clock beyond all
deadlines, and verifies that listings, attempts, replay, idempotency, and outcome
state return to zero. The candidate also measures managed-memory and process
handle deltas after full collection. Failure is any retained state, more than
64 MiB retained managed memory, more than eight retained handles, a working set
above 1.5 GiB, an untyped capacity result, or failure to admit work after expiry.
This accelerated soak intentionally executes far more state lifecycle/cleanup
events than wall-clock traffic would permit. It catches stale deadline-queue
entries, cache growth, replay/idempotency retention, and cleanup cost. Because
it does not open Kestrel/LiteNetLib connections, its process-handle delta is only
a harness guard and is not evidence of transport stability by itself. The
selected production-process gate adds a ten-second real HTTP/UDP transport soak,
samples child-process handles and RSS, asserts bounded growth, then verifies a
clean SIGTERM and socket release. #23 must extend that into the full rate-shaped
multi-client canary while sampling queues, managed memory, and state
cardinalities. A one-hour core override remains required before tagging a
production release.
## Fault and recovery matrix
`run-capacity-gate.sh` runs these deterministic production paths before the
numeric profile:
| Fault | Required result |
| --- | --- |
| HTTP/UDP overload and tracker exhaustion | Typed HTTP `429`/`CapacityExceeded`, silent UDP drop, bounded tracker keys, recovery after the window |
| Optional traffic saturation | Lease-critical renew/update/delete capacity remains available |
| Store/dependency unavailable | Readiness fails; new authorization returns typed `ServiceUnavailable`; liveness remains independent |
| Graceful drain/SIGTERM | New work returns `Draining`; existing pairing may finish; process exits 0 and releases TCP/UDP before the deadline |
| Hard restart | In-flight state is lost; SDK reports typed `ServiceUnavailable`; a host re-registers, rebinds presence, and becomes the only browser-visible replacement |
| UDP listener bind/restart | Readiness stays false without the required listener; rebinding the advertised port restores native LiteNetLib pairing |
| Wall-clock jump/skew | Monotonic lease/attempt authority is neither shortened nor extended; credential skew remains capped at 30 seconds |
| Signing-secret rotation | New key signs, overlap verifies, retired/revoked key rejects, missing material fails startup |
| Principal revocation | Listing, presence, attempts, and outcome paths are removed atomically within the latency budget |
No external database exists in v1, so “dependency/store failure” means the
process-local atomic store is marked unavailable or a required listener/key is
unready. The service fails closed rather than pretending a degraded writable
mode exists.
## Bandwidth and amplification
- Accepted application datagrams are at most 1,200 bytes.
- Malformed, oversized, unauthenticated, stale, replayed, wrong-role, and
rate-limited traffic receives zero response bytes.
- A completing authenticated contribution produces at most one introduction to
each observed peer, and the combined response is at most 2.0 times that
contribution's bytes.
- The frozen-envelope and native LiteNetLib socket tests measure this on the real
UDP listener; the hostile corpus and allocation gate exercise 10,000+ inputs
without input-sized logs, tasks, or queues.
Bandwidth planning must therefore reserve ingress for the configured 2,000
datagrams/second plus edge overhead and egress for a worst-case verified 2.0
amplification. Actual successful pairs normally use two contributions and two
introductions; normal gameplay leaves Rendezvous entirely.
## Availability decision
Single-active remains the v1 topology. The measured core profile proves bounded
state and substantial core-path headroom, while public launch capacity remains
conditional on #23. The service has a bounded stop-before-start restart path.
Its failure domain is deliberately
one process/node/public UDP endpoint: node, kernel, host network, DNS/TLS edge,
secret configuration, or operator error can remove all readiness until the cold
replacement owns the same source-preserving endpoint.
The 99.5% objective permits about 216 minutes of unannounced downtime in a
30-day month. Operations must target process readiness within 15 seconds and
host-visible re-registration within 90 seconds, page when no ready instance
exists, and include detection plus recovery in the monthly budget. The current
in-process test validates typed downtime, same-port HTTP restart, fresh
registration, presence rebinding, and browser visibility in under five seconds;
the production-process test separately validates graceful termination, TCP/UDP
release, replacement startup on the same endpoints, UDP readiness, and the
15-second process-ready RTO. Cold-standby activation policy and the 90-second
operator-to-host recovery objective still require a deployment drill before
release. Rollout and rollback use the
deployment runbook's drain, stop, socket-release, start, smoke sequence; never
overlap old and new active processes.
Bring shared TTL/CAS state and deterministic mediator routing forward before
enabling two active instances if any of these occurs:
- one node cannot sustain 150% of the measured 30-day peak while meeting SLOs;
- CPU stays above 70%, memory above 75%, attempt depth above 70%, or limiter
drops/latency remain elevated after abusive traffic is excluded;
- the availability target rises above 99.5% or planned maintenance must preserve
listings; or
- one region requires multiple simultaneously active mediator endpoints.
Rendezvous makes no multi-instance claim today, so a two-node atomic-pairing
test is intentionally not applicable. It becomes a hard release gate with the
shared-state/routing implementation; until then `SingleActiveInstance=false`
fails production startup. Multi-region and relay remain separate evidence-driven
decisions.
+339
View File
@@ -0,0 +1,339 @@
# Incident and change runbooks
Tracking: #20
These runbooks supplement the [signal and operator reference](observability-and-operator-runbook.md).
Every procedure has four explicit gates: detect, contain, recover, and verify.
Record timestamps, the release digest, bounded aggregates, audit fingerprints,
and `X-Rendezvous-Correlation-ID` values. Never copy credentials, capabilities,
connection tickets, signing material, player identity, raw IP addresses,
endpoints, listing metadata, or full request bodies into an incident record.
Operator routes must be reachable only from an allowed management source. Use a
short-lived, least-permission operator credential minted outside Rendezvous.
Pass it to an approved operator client through protected stdin or a secret agent,
not a URL, command argument, environment-wide process launcher, shell trace, or
ticket. All request shapes and responses are defined by the generated
[OpenAPI v1 document](../api/rendezvous-v1.json).
Before an incident, keep these protected records available without depending on
the affected service: current and previous image digests, matching configuration,
key IDs and lifecycle windows (not raw key values), the game owner/on-call map,
capacity baselines, collector destinations, and a separately authorized
break-glass operator key. Test management-source allowlisting and credential
permissions at least once per release.
Use the exact versioned action shapes below. Confirmation fields deliberately
repeat the target so a stale UI selection or copy error fails closed. Responses
do not echo targets.
| Operation | JSON body |
| --- | --- |
| `POST /v1/operator/listings/revoke` | `{"listingId":"<uuid>","confirmListingId":"<same uuid>"}` |
| `POST /v1/operator/principals/revoke` | `{"subject":"<exact subject>","confirmSubject":"<same subject>","lifetimeSeconds":60}` |
| `POST /v1/operator/keys/revoke` | `{"keyId":"<key id>","confirmKeyId":"<same key id>"}` |
| `POST /v1/operator/drain` | `{"confirmation":"DRAIN"}` |
## Abuse or authentication spike
### Detect
- Alert on a baseline-relative increase in `rendezvous.limiter.drops`, HTTP/UDP
request rate, `rendezvous.operator.authentication` rejected/forbidden results,
registration requests by authentication status, queue depth, or p95/p99 latency.
- Check `/health/live`, `/health/ready`, `rendezvous.store.available`, and
authenticated `GET /v1/operator/status`. Separate public-source rejection,
publisher credential failure, operator probing, and ordinary capacity growth.
- Use only bounded operation/result dimensions and correlation IDs. Do not group
by raw address, token, subject, listing ID, or metadata.
### Contain
- Preserve the dedicated operator partition. Do not raise public limits during
an active spike. Apply source-preserving edge rate controls only when their
collateral effect is understood and UDP source address/port remains intact.
- For one abusive session, call `POST /v1/operator/listings/revoke` with identical
`listingId` and `confirmListingId`. For a confirmed publisher subject, call
`POST /v1/operator/principals/revoke` with identical `subject` and
`confirmSubject` and a 1600 second lifetime.
- Revoke a signing key only when compromise evidence implicates that issuer;
broad key revocation invalidates every credential signed by it. Drain only if
the process itself must be isolated.
### Recover
- Correct the source integration, edge rule, leaked principal grant, or tenant
budget under change control. Let a bounded principal revocation expire only
after the owner confirms remediation; a repeated shorter revocation never
shortens the original deadline.
- Restore normal limits gradually. If saturation caused state churn, allow leases
and attempts to expire naturally rather than deleting arbitrary state.
### Verify
- Require limiter drops, authentication result ratios, queue depth, latency, and
direct-connect outcomes to return to the same-region baseline for the agreed
observation window.
- Confirm readiness stayed healthy or recovered, operator audit contains the
intended action/result fingerprint, revoked resources cannot create new work,
and unaffected tenants can still publish, browse, and connect.
## Signing key or issuer compromise
### Detect
- Treat secret-manager access alerts, unexpected issuance, credentials outside
the expected region/kind, a signing-key expiry alarm, or unexplained publisher
authentication growth as compromise until disproved.
- Identify the non-secret key ID, allowed credential kinds, game/environment
binding, `NotBefore`, `SignUntil`, and `VerifyUntil`. Do not retrieve or paste
raw material merely to compare it.
### Contain
- Stop the affected external issuer and deny further access to its secret.
- From a separate uncompromised break-glass operator key with `RotateKeys`, call
`POST /v1/operator/keys/revoke` with identical `keyId` and `confirmKeyId`.
Runtime revocation is immediate but process-local.
- Remove or mark the key revoked in authoritative provisioning before any
restart. Revoke affected principals/listings when narrower evidence supports
it. Do not drain automatically unless the running instance cannot be trusted.
### Recover
- Generate replacement material in the approved secret boundary, use a new key
ID, bind it to the exact credential kind and tenant, and deploy configuration
referencing the secret—never the secret value.
- Resume issuance with short lifetimes. Reissue only to authenticated workloads.
When confidentiality is lost, do not use normal overlap to keep compromised
credentials valid; document the intentional invalidation window.
- Rotate any release, registry, or operator credential exposed by the same
incident through its owning system; Rendezvous key revocation cannot revoke
unrelated systems.
### Verify
- Confirm `GET /v1/operator/status` shows the compromised key revoked and the
replacement signing, old credentials fail, new exact-scope credentials work,
and the result survives a controlled restart from updated provisioning.
- Pass TestClient registration, browse, authenticated mediation, and direct
traffic with the replacement; monitor authentication and audit results through
at least the maximum newly issued credential lifetime.
## Targeted listing or publisher revocation
### Detect
- Validate the abuse report against game-owned records and bounded Rendezvous
evidence. Determine whether the target is one listing or an authenticated
publisher subject. Do not use display name, metadata, or a raw address as
identity.
- Confirm current aggregate state through `GET /v1/operator/status` and record
the correlation IDs that justified action.
### Contain
- Revoke one listing with `POST /v1/operator/listings/revoke`; the exact listing
UUID must appear in both confirmation fields.
- Revoke a publisher with `POST /v1/operator/principals/revoke`; the exact subject
must appear in both confirmation fields and `lifetimeSeconds` must be 1600.
This removes that principal's active listings and attempts and blocks new ones
for the bounded lifetime.
- Choose the narrowest action. Do not revoke a tenant key for a single listing.
### Recover
- The game owner resolves the ban, account, workload, or configuration issue in
the authoritative game system. Rendezvous does not own user accounts or bans.
- After the original revocation deadline, permit a newly authenticated publisher
to register. There is no un-revoke endpoint and no recovery of removed
ephemeral listings; the host creates a new listing.
### Verify
- Confirm the old listing is no longer browsable or joinable, the principal
cannot publish during its lifetime, and the audit action/result is present
without the raw target.
- Confirm unrelated publishers in the same tenant and another tenant still pass
publish/browse/join/direct-traffic checks.
## Planned restart or crash recovery
### Detect
- Planned restart begins with a recorded change and a healthy current baseline.
Crash recovery begins when liveness/process state fails or both TCP 8080 and
UDP 9050 stop answering. Distinguish dependency/readiness failure from a dead
process; liveness deliberately remains healthy for some recoverable failures.
- Record active listing/lease/attempt aggregates. They are informational only:
v1 has no durable runtime database to restore.
### Contain
- For a planned stop, call `POST /v1/operator/drain` with confirmation exactly
`DRAIN`. Require readiness `503`, liveness `200`, and removal from new traffic.
Allow the bounded drain deadline to finish, then send SIGTERM.
- Never start a second active instance while the old process owns the advertised
HTTP/UDP endpoints. On crash, fence the old process/host and verify both sockets
are released before replacement.
### Recover
- Start exactly one instance from the recorded immutable image digest and matching
reviewed configuration/key references. A restart intentionally loses listings,
observed endpoints, attempts, replay markers, and runtime-only revocations.
- Ensure any emergency key revocation is also present in authoritative
provisioning. Hosts must re-register; clients must browse and start new
attempts. Do not restore stale ephemeral state from logs or backups.
### Verify
- Require live and ready health, UDP bind, store availability, and one active
target. Run the full deployment smoke and confirm host re-registration begins.
- Verify no pre-restart listing or capability is accepted, runtime revocations
that should persist are configuration-backed, and latency/outcomes stabilize.
## Release rollback
### Detect
- Trigger rollback from a predeclared objective: readiness loss, failed deployment
smoke, contract/package incompatibility, security regression, direct-success
regression beyond threshold, or sustained resource regression. Record the new
and previous digests and the evidence; do not move a tag.
### Contain
- Stop promotion and new rollout work. Drain and stop the faulty single active
instance, then verify both public sockets are released. Revoke affected keys or
principals only when the defect creates an authorization risk.
- Preserve logs, artifacts, provenance, signatures, and the faulty release record.
Never overwrite or delete an immutable package/image to reuse its version.
### Recover
- Deploy the previous known-good image by digest with its compatible configuration
and key set. Do not run old and new concurrently. If configuration changed,
apply its reviewed down-migration before starting.
- Publish a corrected build under a new SemVer after diagnosis; mark faulty release
notes withdrawn when appropriate.
### Verify
- Check the running image digest, live/ready health, one active target, UDP source
preservation, and the complete TestClient deployment smoke.
- Confirm package/server compatibility from `GET /v1/operator/status`, hosts
re-register, and the rollback objective returns to baseline for the observation
window.
## Capacity saturation
### Detect
- Page when `rendezvous.queue.depth` remains above 90% of the configured attempt
limit, lease-critical work is shed, `rendezvous.store.available` is zero, or no
ready instance remains. Warn at 70%, sustained `rendezvous.limiter.drops`, or
p95 latency above objective.
- Compare CPU, memory, file descriptors, UDP errors, expiry churn, HTTP operation
rate, and typed connection outcomes with the measured
[capacity profile](capacity-and-resilience.md). Distinguish legitimate growth,
attack traffic, downstream telemetry pressure, and a regression.
### Contain
- Preserve lease-critical and operator reserves. Shed new browse/join work with
the existing typed `429`/`Retry-After` behavior; do not add an unbounded queue.
- Apply per-tenant/source controls at the appropriate trusted boundary. If the
process is unstable, drain new work and recover on one replacement rather than
adding a second active replica; v1 state is process-local.
### Recover
- Remove the causal load or deploy a tested higher single-instance resource and
budget profile. Change CPU/memory and server limits together, using the numeric
gate and accelerated soak before production.
- Long-term horizontal scaling requires a designed shared directory, replay, and
attempt authority. A generic load balancer is not that design.
### Verify
- Re-run the capacity/resilience gate at the chosen profile, then require queue,
limiter drops, expiry churn, latency, store health, and direct-success ratio to
remain within objectives through the production observation window.
- Confirm termination still completes within `DrainDeadlineSeconds + 5` and the
public and operator partitions behave independently.
## Privacy or telemetry incident
### Detect
- Trigger on any credential, token, capability, player identity, raw IP/endpoint,
listing ID, metadata, or caller-reported exact connection duration tied to an
event or identity found in logs, metrics, traces, crash reports, support systems,
or analytics. Aggregate HTTP/UDP duration histograms with bounded operation tags
are expected telemetry. Also trigger when audit data exceeds its approved 30-day
retention without an incident hold.
- Identify the producing version, sink, access population, retention/replication
path, and time window without copying the exposed value into a new system.
### Contain
- Stop or filter the offending export and restrict access to affected sinks.
Preserve the minimum evidence under the incident process; do not take broad
diagnostic dumps that amplify exposure.
- Revoke exposed reusable credentials/keys through their owning boundary. Listing
IDs and endpoints are not authentication secrets, but remove affected listings
if continued exposure creates risk. Notify privacy/security owners according to
applicable policy and law.
### Recover
- Patch the producer to the allowlisted telemetry model, test canary redaction
across logs/metrics/traces/output, and deploy through the immutable release
path. Delete or age out affected data from every sink according to approved
retention and legal-hold direction.
- Replace exposed credentials and re-register hosts when necessary. Do not claim
that a service restart deletes copies already exported to collectors.
### Verify
- Search new telemetry using non-secret synthetic canaries and confirm no canary
or prohibited field crosses the boundary. Verify audit records contain only
fixed fields and fingerprints and that retention/eviction is operating.
- Security/privacy owners confirm sink cleanup, access review, notification, and
monitoring closure before the incident is resolved.
## Dependency or base-image upgrade
### Detect
- Open a reviewed change for an advisory, end-of-support date, pinned-digest
refresh, or planned package update. Record affected package/image, current and
proposed exact version/digest, advisory severity, exploitability, and required
deadline. Never float to `latest` as remediation.
### Contain
- For an actively exploited critical issue, restrict exposure or stop the service
under incident authority while building the fix. Revoking publisher keys does
not repair a vulnerable runtime. Otherwise keep the known-good release running
while the candidate is tested.
### Recover
- Update the SDK/base-image digest, lock files, license/advisory evidence, SBOM,
compatibility matrix, and release notes together. For LiteNetLib or a wire/API
change, apply the explicit version/migration policy rather than silently
replacing compatible bytes.
- Run locked restore, formatting, Debug and Release builds/tests, public contract
and package gates, real consumer restores, reproducible artifact/image builds,
vulnerability scan, signatures, topology/deployment smoke, and capacity checks
proportional to the change. Promote the exact tested digest.
### Verify
- Verify signatures, provenance, checksums, SBOM contents, running digest, and
absence of the advisory in the shipped artifact—not merely the build host.
- Require live/ready health, TestClient direct traffic, real consumer compatibility,
and normal latency/outcomes. Keep the previous digest and compatible config for
rollback until the observation window closes.
@@ -0,0 +1,148 @@
# Observability and operator reference
This runbook defines the production signals and privileged controls for the
Rendezvous service. The service emits `System.Diagnostics.Metrics` instruments
from the `FinalFactory.Rendezvous` meter and distributed-tracing activities from
`FinalFactory.Rendezvous.Server`. Connect those sources to the deployment's
OpenTelemetry or equivalent collector. Do not add identifiers to metric labels.
Concrete detect/contain/recover/verify procedures for abuse, key compromise,
targeted revocation, restart, rollback, saturation, privacy incidents, and
dependency upgrades are in the [incident and change runbooks](incident-runbooks.md).
## Health and readiness
- `GET /health/live` proves that the HTTP process can answer. It deliberately
remains independent of provisioning, the state store, drain state, and optional
listeners so an orchestrator does not restart a recoverable dependency failure.
- `GET /health/ready` returns success only after the HTTP path is answering, the
required IPv4 UDP socket is bound, any configured IPv6 UDP socket is bound,
provisioning loaded successfully, the store is available, and drain has not
started. A failed check returns `503` and removes the instance from new work.
- A graceful drain immediately makes readiness fail while liveness remains healthy.
Existing work may complete until the bounded store drain deadline.
## Metrics and traces
| Instrument | Purpose | Bounded dimensions |
| --- | --- | --- |
| `rendezvous.http.requests` / `rendezvous.http.duration` | HTTP volume and latency | operation, status code |
| `rendezvous.udp.results` / `rendezvous.udp.duration` | UDP mediation volume and processing latency | frozen/litenet operation, result |
| `rendezvous.limiter.drops` | Requests shed by admission controls | transport, fixed partition class |
| `rendezvous.operator.authentication` | Accepted, forbidden, and rejected operator authentication | result |
| `rendezvous.audit.events` | Privileged action outcomes | fixed action, result |
| `rendezvous.connection.outcomes` | Client-reported direct-connect outcomes | normalized outcome, elapsed bucket |
| `rendezvous.pairing.latency` | Time from attempt creation to successful peer introduction | none |
| `rendezvous.queue.depth` | Active join-attempt queue depth | none |
| `rendezvous.store.active_listings` / `active_leases` / `active_attempts` / `replay_markers` | Current ephemeral load | none |
| `rendezvous.store.expiry_churn` | Cumulative natural expiry activity | none |
| `rendezvous.store.available` | Store health (`1` available, `0` unavailable) | none |
HTTP responses include `X-Rendezvous-Correlation-ID`. It is a generated trace ID
or random value, never a caller-supplied session or player identifier. UDP and
HTTP activities contain operation-level data only. Logs and traces must not add
tokens, capabilities, session/listing IDs, player subjects, metadata, raw IP
addresses, or endpoint values.
Recommended dashboard panels are request rate and p50/p95/p99 latency by fixed
operation, UDP result ratio, direct connection success ratio, pairing latency,
active listings/attempts, expiry churn, limiter drops, store availability,
operator authentication results, audit action results, and signing-key windows.
## Alerts
Tune thresholds from the normal production baseline, then keep these conditions
as distinct actionable alerts:
- **Signing key expiry:** page when any required signing key has less than seven
days before `signUntil`; escalate at 24 hours. Confirm a replacement is signing
and the previous key remains verify-only for the maximum credential lifetime.
- **Authentication spike:** warn when rejected or forbidden operator authentication
exceeds five attempts in five minutes. Treat unexpected publisher-authentication
growth as a possible credential or integration incident.
- **Direct success regression:** warn when the connected outcome ratio falls more
than 20% below its seven-day same-region baseline for 15 minutes, with a minimum
sample floor. Break down only by bounded outcome and time bucket.
- **Saturation:** warn when queue depth remains above 70% of the configured attempt
limit, limiter drops are sustained, or p95 latency exceeds the service objective;
page at 90% or when lease-critical traffic is shed.
- **Store degradation:** page immediately when `rendezvous.store.available` is zero
or readiness fails for the store. Rising expiry churn without corresponding new
work is a warning for stalled clients or clock/configuration mistakes.
- **Listener/config readiness:** page when no ready instances remain. Investigate
UDP bind failures, a configured-but-unbound IPv6 listener, provisioning errors,
and unintended drain state separately.
## Operator authentication and controls
Operator credentials use a signing key configured with `CredentialKinds:
["Operator"]`. Operator keys cannot be scoped to a game/environment or used for
publisher credentials. Mint short-lived operator credentials through the trusted
provisioning process, outside the public Rendezvous HTTP service, and grant only
the required permission. Never place credentials in command history, URLs, logs,
or support tickets.
The application also enforces a default-deny source boundary. Configure at most
32 exact operator source IPs in
`Rendezvous:AbuseProtection:OperatorAllowedAddresses`; an empty list disables all
operator HTTP access. Development permits loopback only. Production must place
`/v1/operator/*` behind a private management listener or reverse-proxy ACL, list
only the resulting trusted management source addresses, and block that path on
the public edge. If forwarded headers are enabled, keep the existing exact-proxy,
single-hop trust policy and allowlist the post-forwarding operator source. Verify
from both an allowed management host and a denied public host before deployment.
Denied sources are charged to the bounded general HTTP partition before credential
or request-body processing, then receive `404`; sustained denied traffic receives
the same typed `429` overload response as other public traffic.
Operator traffic has a dedicated, bounded rate/concurrency partition and critical
tracker-key reserve. Public browse/join saturation therefore cannot consume the
operator control budget, while compromised management sources remain rate-limited.
The OpenAPI document defines the separate `OperatorBearer` scheme. All endpoints
are under `/v1/operator`:
| Endpoint | Permission | Confirmation |
| --- | --- | --- |
| `GET /status` | `ReadPolicy` | none; returns aggregates, tenant status, safe key status, and audit counts |
| `POST /listings/revoke` | `RevokePublisher` | repeat the exact listing ID in `confirmListingId` |
| `POST /principals/revoke` | `RevokePublisher` | repeat the exact subject and choose a 1-600 second revocation lifetime |
| `POST /keys/revoke` | `RotateKeys` | repeat the exact key ID; runtime revocation is immediate |
| `POST /drain` | `ManagePolicy` | send the exact value `DRAIN` |
Publisher credentials are rejected on this surface even if their subject resembles
an operator. Destructive responses do not echo identifiers. The status response
does not expose player identities, raw endpoints, session metadata, capabilities,
or tokens. Every authenticated operator action, rejected confirmation, and
permission denial is audited with actor and target fingerprints.
Key revocation is process-local in the current single-instance store. Apply the
same revocation to every instance, then replace configuration before restarting;
a restart reconstructs the configured key ring. Principal revocation is bounded
to ten minutes and removes that principal's active listings and attempts. A
repeat action may extend an active revocation but never shortens it; wait for its
original deadline rather than treating a shorter repeat as an un-revoke. Use
listing revocation for one targeted session and drain before planned shutdown.
## Audit retention and incident handling
The in-process audit trail defaults to 10,000 entries and 30 days. It evicts the
oldest record at capacity and purges expired records on the next write. Configure
`Rendezvous:Audit:MaxEntries` and `RetentionDays` within their validated bounds.
Export the structured `AuditTrail` log events through the deployment's protected
logging pipeline when durable retention is required; the in-memory trail is not a
durable compliance archive. Those events include only timestamps, fixed action
fields, correlation IDs, and actor/target fingerprints.
Audit records retain timestamp, fixed action/result, target kind, correlation ID,
and 96-bit SHA-256 fingerprints of actor and target. Routine logs contain only the
fixed action/result/target kind and correlation ID. Restrict audit access to the
operator role, retain aggregates only as long as operationally necessary, and
delete raw exported audit data according to the 30-day policy unless an incident
hold is approved.
During an incident: confirm readiness and store health; capture aggregate graphs
and correlation IDs; revoke the narrowest listing, principal, or key; drain only
when isolation is required; record the action in the incident timeline; and verify
that direct success, limiter drops, and authentication rates return to baseline.
Do not copy player data, endpoints, or credentials into the incident record.
+150
View File
@@ -0,0 +1,150 @@
# Releases and compatibility
Tracking: #19
Rendezvous releases are immutable, reproducible, and promoted only after the
same candidate has passed package, consumer, server, container, and staging
checks. A release consists of matching Client and Contracts NuGet packages, a
framework-dependent Linux server archive, a versioned linux/amd64 OCI image,
package/runtime and container SPDX inventories, checksums, provenance, release
notes, and signatures. No workflow publishes a `latest` tag.
## Version dimensions
The central values in `eng/Versions.props` are the authority. Client,
Contracts, and Server use SemVer. HTTP, UDP mediation, and connection-ticket
formats advance independently so a wire change cannot hide inside a package
patch release. The current machine-readable matrix is
[`compatibility.json`](compatibility.json); the authenticated operator status
endpoint exposes the server's supported window at runtime.
| Surface | Current | Compatibility rule |
| --- | ---: | --- |
| Client and Contracts | 1.0.0 | Matching exact versions; source/API breaks require a package major bump. |
| Server | 1.0.0 | Accepts Client 1.0.0 through compatible 1.x releases. |
| HTTP contract | 1 | Frozen OpenAPI, JSON vectors, and public API snapshot. |
| UDP mediation | 1 | Frozen codec vectors; incompatible bytes require UDP v2. |
| Connection ticket | 1 | A format change requires a new accepted ticket version and migration window. |
| LiteNetLib | 2.1.4 | Exact dependency; LiteNetLib 1.x is rejected by package and consumer gates. |
| Gameplay protocol | Per tenant | Exact match; Rendezvous does not translate gameplay protocols. |
`scripts/check-compatibility.sh` compares protected snapshots against the base
revision. A changed public API snapshot requires a package major increase; an
HTTP or UDP golden surface requires the corresponding contract increase. The
normal tests also compare implementation output with the current versioned
snapshots. For a deliberate break, add a new versioned contract directory and
documentation instead of replacing the prior version's evidence.
## Candidate build
From a clean tagged checkout:
```bash
./scripts/check-release-tag.sh v1.0.0
./scripts/build-release.sh 1.0.0
./scripts/verify-release.sh 1.0.0
```
The tag build runs inside the digest-pinned `release-builder` Docker stage,
which combines the pinned SDK with a pinned Python runtime. It uses the locked
dependency graph, enforces NuGet
advisories and approved licenses, runs formatting/build/tests, regenerates the
OpenAPI drift check, and packs twice after a clean rebuild. NuGet's random OPC
relationship identifiers are canonicalized before comparison; both `.nupkg`
and `.snupkg` outputs must then be byte-identical. Package metadata identifies the exact
repository commit, portable PDBs carry SourceLink data, and the Linux archive,
runtime SBOM timestamp, and checksum ordering are deterministic. Buildx and
BuildKit are also pinned for the linux/amd64 OCI build. Provenance records each
artifact-producing tool version; an out-of-band rebuild must use the pinned
builder and recorded versions rather than treating the runner label as a
reproducibility guarantee.
All project-authored artifact normalization and checksum updates run inside the
same pinned builder; host tools only orchestrate or verify. The separately
pinned Trivy and Cosign tools produce the container inventory and signatures.
The tag workflow also performs two no-cache image builds with the commit time
and revision fixed, disables unsigned builder-generated attestations, and
requires identical OCI image IDs before signing the project provenance.
The local gate builds net8.0 SpaceGame- and Unscouted-shaped API fixtures using
only the candidate feed plus NuGet.org; the Unscouted fixture also carries its
real direct LiteNetLib 2.1.4 pin. The tag gate separately checks out the exact
SpaceGame and Unscouted revisions in `eng/consumer-revisions.json`, injects
exact candidate references without modifying those repositories, and restores
their real game/network projects. Both paths must resolve the matching Client
and Contracts version and LiteNetLib 2.1.4. Updating a consumer revision is a
reviewed compatibility change, not a floating-main check.
## Promotion and publication
Pushing the matching `vMAJOR.MINOR.PATCH` tag starts the tag-only release
workflow. Before any external write it:
1. builds and verifies the artifact set;
2. builds the exact versioned container candidate;
3. starts that image with production hardening and a temporary staging key;
4. completes HTTP health, registration, browse, authenticated UDP mediation,
and direct traffic;
5. rejects all high or critical container findings and emits a container SPDX
inventory;
6. finalizes and verifies checksums over the publish-ready artifact set; and
7. confirms the two NuGet versions, container version, and Gitea release do not
already exist.
Publication has no skip-duplicate behavior. Gitea's immutable package versions,
the workflow concurrency lock, and the preflight make a successful tag a
single publication event. The workflow pushes symbols, publishes only the
versioned container tag, records its `sha256` digest in both a digest file and
provenance, regenerates the checksum manifest so that digest and public key are
covered, signs the checksum manifest and image, attaches signed provenance,
verifies the complete published-set schema and all signatures, and creates the
Gitea release with exactly those artifacts. The detached checksum signature
bundle is the sole envelope excluded from its own signed manifest.
The loaded image ID is captured immediately after the byte-reproducible build;
publication refuses to push if staging or another process retagged that local
name to different bytes.
The protected `production` environment requires these secrets:
- `RELEASE_TOKEN`: a dedicated Gitea token limited to this repository and the
HeiKyu package registry, with repository and package write access;
- `RELEASE_USERNAME`: the dedicated Gitea service-account name that owns the
release token;
- `COSIGN_PRIVATE_KEY`: the encrypted Cosign release private key; and
- `COSIGN_PASSWORD`: its password, stored separately.
No development signing key, registry credential, or deployable configuration
is stored in source or packages. Keep the Cosign public key with operational
records. Rotate the release key between releases: retain the old public key for
historical verification, install the new encrypted private key and password as
one reviewed change, verify a signed non-release blob, and only then retire the
old secret. Suspected compromise requires token/key revocation and a new
version; never overwrite or delete evidence to reuse a released version.
## Release notes and migration
Every dated `CHANGELOG.md` entry must contain Compatibility, Security and
configuration, and Migration sections. Before tagging, state the supported
Client/server window, all HTTP/UDP/ticket changes, security fixes, required
configuration, and operator/consumer migration steps.
For a protocol migration, first make the server read both old and new versions
within an explicit bounded window, publish a Client that writes the new version,
verify adoption through bounded telemetry, then remove the old reader only in a
subsequent breaking release. Never silently reinterpret old bytes. Consumers
pin both Rendezvous packages to one exact version and choose gameplay protocol
compatibility per tenant.
## Rollback and interrupted publication
Runtime rollback means redeploying the previous known-good image by digest and
its matching configuration; it does not move a tag. Packages and release
records remain available so already restored clients stay reproducible. If a
new release is faulty, revoke affected publisher or signing keys when relevant,
mark the release notes as withdrawn, and publish the fix under a new SemVer.
The registries cannot provide a transaction spanning NuGet, OCI, signatures,
and release attachments. If publication stops after its first external write,
the preflight intentionally prevents an automatic rerun. An operator must
inventory every destination, preserve logs and hashes, complete or withdraw the
partial version under change control, and then issue a new version. This avoids
turning a partial failure into an untraceable overwrite.
+27
View File
@@ -0,0 +1,27 @@
{
"schemaVersion": 1,
"release": "1.0.0",
"server": {
"minimumClientVersion": "1.0.0",
"maximumClientMajorVersion": 1
},
"packages": {
"FinalFactory.Rendezvous.Client": "1.0.0",
"FinalFactory.Rendezvous.Contracts": "1.0.0"
},
"contracts": {
"http": [1],
"udp": [1],
"connectionTicket": [1],
"gameplay": "exact-per-tenant"
},
"transport": {
"package": "LiteNetLib",
"version": "2.1.4",
"major": 2
},
"consumers": {
"SpaceGame": "net8.0",
"Unscouted": "net8.0"
}
}
+103
View File
@@ -0,0 +1,103 @@
# Hostile-input and overload protection
Tracking: #15
Rendezvous treats every public HTTP request and UDP datagram as hostile. The
server applies bounded fixed-window request budgets and concurrency ceilings in
two stages so malformed input is discarded before expensive work while valid
traffic is also isolated by its authenticated scope.
## Enforcement order
1. Kestrel and the HTTP abuse middleware cap request bodies at 16 KiB. A known
oversized body receives a typed `413` response before endpoint dispatch.
2. Every HTTP request consumes global, source-prefix, and operation budgets and
acquires the corresponding concurrency leases. IPv4 sources share a `/24`
budget and IPv6 sources share a `/56` budget; raw addresses are not retained.
Non-lease operations also consume a smaller optional-work budget, leaving a
configured global and source-prefix reserve for renew, update, and delete
operations during shedding.
Health probes use their own source-prefix budget so public API overload cannot
make a healthy instance fail its orchestrator probes, while health traffic is
still bounded.
Operator endpoints likewise use a separate bounded rate/concurrency partition
backed by the critical tracker reserve. They first require an exact source IP
from the default-deny `OperatorAllowedAddresses` policy, so public traffic
cannot spend the incident-response budget.
3. Once an endpoint has safely derived identities, it also acquires applicable
tenant, principal or capability, and listing/attempt budgets. Secret
capabilities are represented only by bounded SHA-256 fingerprints.
4. Every UDP envelope consumes global, source-prefix, and wire-operation
budgets before decoding. A structurally and cryptographically valid request
then consumes capability, role, and mediation-handle budgets before state
mutation or introduction.
5. HTTP overload returns the stable `RateLimited` error, status `429`, and a
bounded `Retry-After` value in both the header and response contract. UDP
overload and every invalid UDP input are silently dropped.
The same HTTP identity budget is computed whether or not a listing or attempt
exists. Rejection therefore does not disclose resource existence. Publisher
authentication also completes before any tenant/resource operation, while the
pre-authentication source budget prevents invalid credentials from bypassing
load shedding.
## Bounded state and recovery
`Rendezvous:AbuseProtection:MaxTrackedKeys` is a hard combined ceiling for rate
and active-concurrency keys. General HTTP and UDP traffic cannot consume the
configured `CriticalTrackedKeyReserve`; lease operations, health probes, and
allowlisted operator controls may
use that reserve but never exceed the hard ceiling. A request that would exceed
its applicable ceiling fails closed without adding state. Fixed-window rate keys
are cleared at the next window boundary; concurrency keys are removed as their
request leases finish. HTTP and UDP trackers have separate locks and cardinality
partitions, so a UDP flood cannot block HTTP admission on a shared lock or
consume HTTP key capacity. This gives
deterministic burst recovery and prevents an attacker from growing a permanent
high-cardinality address, credential, or resource table.
The complete default profile is checked into
`src/FinalFactory.Rendezvous.Server/appsettings.json`. Operators may lower or
tune limits for a measured deployment profile, but must preserve all dimensions
and leave the tracker ceiling above the maximum simultaneous key set. A rolling
deployment should use the same profile on every instance. These per-process
limits are a final service boundary; an edge proxy may add stricter distributed
limits but is not a substitute for them.
When an HTTP reverse proxy is used, every immediate proxy address must be
allowlisted in `Rendezvous:AbuseProtection:TrustedProxyAddresses` (or indexed
environment variables such as
`Rendezvous__AbuseProtection__TrustedProxyAddresses__0`). Only one forwarded
hop is accepted. With an empty allowlist, forwarded headers are ignored and the
direct TCP peer is the source. Never add a broad network range or accept
untrusted `X-Forwarded-For` input: that would let a caller choose its own rate
partition.
## Reflection, disclosure, and logging rules
- UDP sends nothing for malformed, oversized, unauthenticated, stale,
replayed, wrong-role, or rate-limited input.
- Introductions are emitted only after both role-scoped capabilities bind to
their observed gameplay-socket sources. HTTP never supplies a public
introduction target.
- Private candidates must be same-family private unicast addresses and are used
only for peers observed behind the same public address.
- Abuse keys, exceptions, and responses never include bearer credentials,
capabilities, tickets, raw endpoints, metadata values, or hostile markup.
- Endpoint and capability values are not used as metric labels or log fields.
## Verification
The deterministic test corpora use the recorded seeds `0x152026`, `0x154A50`,
and `0x1557A7E`. They exercise 10,000 arbitrary UDP envelopes through the
production decoder, 5,000 arbitrary HTTP/credential parser inputs, and 1,000
mutated state transitions, including the oversized and configured-capacity
boundaries.
Focused tests cover IPv4 and IPv6 prefix
partitioning, tenant/principal/resource concurrency, tracker exhaustion,
window recovery, wire-operation isolation, a steady-state allocation ceiling,
typed `429`/`413` responses, secret fingerprint redaction, and silent
authenticated UDP shedding. The existing state, contract, HTTP, client,
and mediator suites continue to cover cross-tenant access, replay, role swaps,
credential rotation, bounded metadata, endpoint validation, and one-shot
amplification behavior.
+1 -1
View File
@@ -11,7 +11,7 @@ backlog where the control is implemented and verified.
| Per-game credentials and signing keys | Provisioned principals and versioned keys are scoped to game/environment; secrets come from a provider and never a public binary. (#5) | Cross-tenant authorization tests, rotation/overlap/revocation tests, and secret scans. |
| Short-lived, single-purpose tokens resistant to replay | Issuer fixes audience, tenant, attempt, role, issued/expiry times, nonce, and key ID; store atomically consumes nonce/ticket. (#4, #6, #10) | Golden vectors; expired, future, mutated, wrong-role, wrong-tenant, and concurrent replay tests. |
| Strict payload, metadata, and token size limits | ADR 0003 ceilings are checked before allocation/deserialization and again at domain construction. (#4, #15) | Boundary/property tests, malformed corpus, and allocation-aware fuzzing. |
| Registration, query, and introduction rate limits | Layered per-address, principal, tenant, and global token buckets with bounded queues and stable retry guidance. (#15) | Limit partition/isolation tests and overload/soak profiles. |
| Registration, query, and introduction rate limits | Layered fixed-window budgets and concurrency leases cover global, operation, IPv4 `/24` or IPv6 `/56`, tenant, principal/capability, and listing/attempt dimensions with a bounded key table and stable retry guidance. (#15) | Deterministic partition, concurrency, tracker-exhaustion, recovery, typed-overload, and silent-UDP-shedding tests. |
| Lease expiry removes abandoned servers | Visibility and join eligibility atomically require a fresh lease and fresh authenticated presence. (#6, #7) | Fake-clock expiry, renew/expire race, restart, and stale-host join tests. |
| Validate game, environment, room, and protocol boundaries | Every identifier is a validated type; store keys and authorization decisions include server-derived tenant scope; protocol is exact-match in v1. (#4-#10) | Contract, tenant-isolation, incompatible-version, and confused-deputy tests. |
| Structured audit events without secrets or reusable credentials | Allowlisted audit schema excludes metadata values, raw endpoints, tokens, and key material; event volume is bounded. (#16) | Captured-log/audit assertions and credential canary scans. |
+7 -4
View File
@@ -63,8 +63,9 @@ only its public key ID/lifecycle metadata and does not require retired secret
material to remain available.
Key IDs are non-secret base64url identifiers. Secret references are resolved
through `ISecretProvider`; production supports `env:<VARIABLE>` references and
the interface is replaceable by a deployment-specific vault/KMS adapter. The
through `ISecretProvider`; production supports base64 `env:<VARIABLE>` and raw
`file:/absolute/path` references to bounded non-symlink files. The interface is
replaceable by a deployment-specific vault/KMS adapter. The
committed development profile uses an in-memory random key identified by a
`development:ephemeral/...` reference. It never writes key material to disk and
all credentials become invalid when the process exits.
@@ -74,8 +75,10 @@ all credentials become invalid when the process exits.
`Rendezvous:Provisioning` supplies issuer, audience, clock skew, signing-key
descriptors, and game policies. A production key reference such as
`env:RENDEZVOUS_SIGNING_KEY_2026_01` expects that environment variable to hold at
least 32 random bytes encoded as base64. Missing, malformed, short, inactive, or
duplicate keys stop startup with a key-ID-only diagnostic. No game-wide secret
least 32 random bytes encoded as base64. `file:/run/secrets/rendezvous-signing`
expects the raw bytes in a read-only, absolute, non-symlink file. Missing,
malformed, short, inactive, or duplicate keys stop startup with a key-ID-only
diagnostic. No game-wide secret
belongs in `appsettings`, source control, examples, the Client package, URLs,
responses, logs, metrics, exceptions, or diagnostic dumps.
+15
View File
@@ -0,0 +1,15 @@
<Project>
<PropertyGroup>
<RendezvousVersion>1.0.0</RendezvousVersion>
<RendezvousMajorVersion>1</RendezvousMajorVersion>
<RendezvousMinorVersion>0</RendezvousMinorVersion>
<RendezvousPatchVersion>0</RendezvousPatchVersion>
<MinimumClientVersion>1.0.0</MinimumClientVersion>
<MaximumClientMajorVersion>1</MaximumClientMajorVersion>
<HttpContractVersion>1</HttpContractVersion>
<UdpContractVersion>1</UdpContractVersion>
<ConnectionTicketFormatVersion>1</ConnectionTicketFormatVersion>
<LiteNetLibVersion>2.1.4</LiteNetLibVersion>
<LiteNetLibMajorVersion>2</LiteNetLibMajorVersion>
</PropertyGroup>
</Project>
+17
View File
@@ -0,0 +1,17 @@
{
"schemaVersion": 1,
"consumers": [
{
"name": "SpaceGame",
"repository": "https://git.finalfactory.de/Kyuubi/SpaceGame.git",
"revision": "77519b0cc418a27f8d408ae2d7b8812fbe087c04",
"project": "SpaceGame.csproj"
},
{
"name": "Unscouted",
"repository": "https://git.finalfactory.de/HeiKyu/Unscouted.git",
"revision": "7807dbee86eb8b98e702f1eb89c88adff728f635",
"project": "Net.Core/Net.Core.csproj"
}
]
}
+14
View File
@@ -0,0 +1,14 @@
# syntax=docker/dockerfile:1.7@sha256:a57df69d0ea827fb7266491f2813635de6f17269be881f696fbfdf2d83dda33e
FROM python:3.12.11-slim-bookworm@sha256:c00fc7b44d844b6da22861ec24af43968a5200eac4ec607b4725d585165d6b49 AS release-python
FROM ghcr.io/jqlang/jq:1.8.1@sha256:95de8f005ca027686a1ca3b0853e2bb219062438015862816159f3f25a4d4230 AS release-jq
FROM mcr.microsoft.com/dotnet/sdk:10.0.301-noble@sha256:ea8bde36c11b6e7eec2656d0e59101d4462f6bd630730f2c8201ed0572b295d5 AS release-builder
COPY --from=release-python /usr/local/ /usr/local/
COPY --from=release-jq /jq /usr/local/bin/jq
RUN dotnet --version \
&& python3 --version \
&& git --version \
&& tar --version \
&& gzip --version \
&& jq --version
WORKDIR /source
+27
View File
@@ -0,0 +1,27 @@
{
"schemaVersion": 1,
"packageRegistry": "https://git.finalfactory.de/api/packages/HeiKyu/nuget/index.json",
"containerRepository": "git.finalfactory.de/heikyu/rendezvous",
"allowedLicenseExpressions": [
"Apache-2.0",
"BSD-2-Clause",
"BSD-3-Clause",
"MIT"
],
"dependencyLicenseOverrides": {
"xunit.abstractions/2.0.3": {
"license": "Apache-2.0",
"reason": "Legacy package predates NuGet SPDX metadata; reviewed against the xUnit.net Apache-2.0 license."
}
},
"publishedPackages": [
"FinalFactory.Rendezvous.Client",
"FinalFactory.Rendezvous.Contracts"
],
"forbiddenArtifactNameFragments": [
"credential",
"password",
"private-key",
"signing-key"
]
}
+823
View File
@@ -0,0 +1,823 @@
#!/usr/bin/env python3
import argparse
import datetime as dt
import hashlib
import json
import os
import pathlib
import re
import sys
import zipfile
import xml.etree.ElementTree as ET
from xml.sax.saxutils import escape
PROJECT_PACKAGE_PREFIX = "finalfactory.rendezvous."
CORE_PROPERTIES_PATH = (
"package/services/metadata/core-properties/core-properties.psmdcp"
)
def fail(message: str) -> None:
raise SystemExit(message)
def load_json(path: pathlib.Path):
with path.open(encoding="utf-8") as stream:
return json.load(stream)
def dependency_inventory(root: pathlib.Path):
dependencies = {}
for lock_path in sorted(root.glob("**/packages.lock.json")):
if any(part in {"bin", "obj", "artifacts"} for part in lock_path.parts):
continue
lock = load_json(lock_path)
for framework in lock.get("dependencies", {}).values():
for package_id, details in framework.items():
resolved = details.get("resolved")
if not resolved or package_id.lower().startswith(PROJECT_PACKAGE_PREFIX):
continue
key = package_id.lower()
previous = dependencies.get(key)
if previous is not None and previous[1] != resolved:
fail(
f"Dependency {package_id} resolves to both {previous[1]} and {resolved}."
)
dependencies[key] = (package_id, resolved)
return [dependencies[key] for key in sorted(dependencies)]
def package_license(package_id: str, version: str, overrides):
package_root = pathlib.Path(
os.environ.get("NUGET_PACKAGES", pathlib.Path.home() / ".nuget" / "packages")
)
version_dir = package_root / package_id.lower() / version
nuspecs = list(version_dir.glob("*.nuspec"))
if len(nuspecs) != 1:
fail(f"Expected one restored nuspec for {package_id} {version} in {version_dir}.")
root = ET.parse(nuspecs[0]).getroot()
license_element = root.find(".//{*}license")
if license_element is None or license_element.get("type") != "expression":
override = overrides.get(f"{package_id.lower()}/{version}")
if override is None:
fail(f"{package_id} {version} does not declare an SPDX license expression.")
return override["license"]
expression = (license_element.text or "").strip()
if not expression:
fail(f"{package_id} {version} has an empty license expression.")
return expression
def command_policy(args) -> None:
root = pathlib.Path(args.root).resolve()
policy = load_json(root / "eng" / "release-policy.json")
allowed = set(policy["allowedLicenseExpressions"])
inventory = dependency_inventory(root)
observed = []
for package_id, version in inventory:
expression = package_license(
package_id, version, policy.get("dependencyLicenseOverrides", {})
)
if expression not in allowed:
fail(
f"Dependency {package_id} {version} uses unapproved license {expression}."
)
observed.append({"id": package_id, "version": version, "license": expression})
lite_net_lib = [item for item in observed if item["id"].lower() == "litenetlib"]
if lite_net_lib != [{"id": "LiteNetLib", "version": "2.1.4", "license": "MIT"}]:
fail(f"LiteNetLib must resolve exactly to the reviewed 2.1.4 release: {lite_net_lib}")
print(json.dumps({"dependencies": observed}, indent=2))
def command_audit(args) -> None:
document = load_json(pathlib.Path(args.input))
findings = []
def visit(value):
if isinstance(value, dict):
if value.get("vulnerabilities"):
findings.append(value)
for child in value.values():
visit(child)
elif isinstance(value, list):
for child in value:
visit(child)
visit(document)
if findings:
fail(f"Locked dependency graph contains known vulnerabilities: {findings}")
print("Locked dependency graph has no reported vulnerabilities.")
def release_dependency_graph(root: pathlib.Path, server_deps: pathlib.Path):
dependencies = {}
edges = set()
server_document = load_json(server_deps)
for package_key, details in server_document.get("libraries", {}).items():
if details.get("type") != "package":
continue
package_id, version = package_key.rsplit("/", 1)
dependencies[package_id.lower()] = (package_id, version)
server_target = next(iter(server_document.get("targets", {}).values()), {})
for source_key, details in server_target.items():
source_id = source_key.rsplit("/", 1)[0]
source = (
"SPDXRef-Server"
if source_id == "FinalFactory.Rendezvous.Server"
else source_id.lower()
)
for dependency_id in details.get("dependencies", {}):
target = {
"FinalFactory.Rendezvous.Contracts": "SPDXRef-Contracts",
"FinalFactory.Rendezvous.Client": "SPDXRef-Client",
}.get(dependency_id, dependency_id.lower())
edges.add((source, target))
lock_roots = (
("SPDXRef-Client", root / "src/FinalFactory.Rendezvous.Client/packages.lock.json"),
(
"SPDXRef-Contracts",
root / "src/FinalFactory.Rendezvous.Contracts/packages.lock.json",
),
)
for root_id, lock_path in lock_roots:
lock = load_json(lock_path)
for framework in lock.get("dependencies", {}).values():
for package_id, details in framework.items():
resolved = details.get("resolved")
if resolved and not package_id.lower().startswith(PROJECT_PACKAGE_PREFIX):
dependencies[package_id.lower()] = (package_id, resolved)
if details.get("type") == "Direct":
edges.add((root_id, package_id.lower()))
source = package_id.lower()
for dependency_id in details.get("dependencies", {}):
if dependency_id.lower().startswith(PROJECT_PACKAGE_PREFIX):
continue
edges.add((source, dependency_id.lower()))
edges.add(("SPDXRef-Client", "SPDXRef-Contracts"))
inventory = [dependencies[key] for key in sorted(dependencies)]
return inventory, edges
def command_sbom(args) -> None:
root = pathlib.Path(args.root).resolve()
policy = load_json(root / "eng" / "release-policy.json")
overrides = policy.get("dependencyLicenseOverrides", {})
packages = [
{
"SPDXID": "SPDXRef-Rendezvous",
"name": "FinalFactory.Rendezvous",
"versionInfo": args.version,
"downloadLocation": "NOASSERTION",
"filesAnalyzed": False,
"licenseConcluded": "NOASSERTION",
"licenseDeclared": "NOASSERTION",
"copyrightText": "NOASSERTION",
}
]
internal_packages = [
("SPDXRef-Server", "FinalFactory.Rendezvous.Server"),
("SPDXRef-Client", "FinalFactory.Rendezvous.Client"),
("SPDXRef-Contracts", "FinalFactory.Rendezvous.Contracts"),
]
for spdx_id, package_id in internal_packages:
packages.append(
{
"SPDXID": spdx_id,
"name": package_id,
"versionInfo": args.version,
"downloadLocation": "NOASSERTION",
"filesAnalyzed": False,
"licenseConcluded": "NOASSERTION",
"licenseDeclared": "NOASSERTION",
"copyrightText": "NOASSERTION",
}
)
inventory, dependency_edges = release_dependency_graph(
root, pathlib.Path(args.server_deps)
)
dependency_ids = {}
for index, (package_id, version) in enumerate(
inventory, start=1
):
expression = package_license(package_id, version, overrides)
spdx_id = f"SPDXRef-Package-{index}"
dependency_ids[package_id.lower()] = spdx_id
packages.append(
{
"SPDXID": spdx_id,
"name": package_id,
"versionInfo": version,
"downloadLocation": "NOASSERTION",
"filesAnalyzed": False,
"licenseConcluded": expression,
"licenseDeclared": expression,
"copyrightText": "NOASSERTION",
"externalRefs": [
{
"referenceCategory": "PACKAGE-MANAGER",
"referenceType": "purl",
"referenceLocator": f"pkg:nuget/{package_id}@{version}",
}
],
}
)
created = dt.datetime.fromtimestamp(
int(args.source_date_epoch), dt.timezone.utc
).strftime("%Y-%m-%dT%H:%M:%SZ")
document = {
"spdxVersion": "SPDX-2.3",
"dataLicense": "CC0-1.0",
"SPDXID": "SPDXRef-DOCUMENT",
"name": f"FinalFactory.Rendezvous-{args.version}",
"documentNamespace": (
"https://git.finalfactory.de/HeiKyu/Rendezvous/sbom/"
f"{args.version}/{args.commit}"
),
"creationInfo": {
"created": created,
"creators": ["Tool: eng/release_artifacts.py"],
},
"documentDescribes": ["SPDXRef-Rendezvous"],
"packages": packages,
"relationships": [
{
"spdxElementId": "SPDXRef-Rendezvous",
"relationshipType": "CONTAINS",
"relatedSpdxElement": spdx_id,
}
for spdx_id, _ in internal_packages
]
+ [
{
"spdxElementId": dependency_ids.get(source, source),
"relationshipType": "DEPENDS_ON",
"relatedSpdxElement": dependency_ids.get(target, target),
}
for source, target in sorted(dependency_edges)
if source in dependency_ids or source.startswith("SPDXRef-")
if target in dependency_ids or target.startswith("SPDXRef-")
],
}
output = pathlib.Path(args.output)
output.parent.mkdir(parents=True, exist_ok=True)
output.write_text(json.dumps(document, indent=2) + "\n", encoding="utf-8")
def nuspec_metadata(archive: pathlib.Path):
with zipfile.ZipFile(archive) as package:
names = package.namelist()
nuspecs = [name for name in names if name.endswith(".nuspec")]
if len(nuspecs) != 1:
fail(f"{archive.name} must contain exactly one nuspec.")
root = ET.fromstring(package.read(nuspecs[0]))
metadata = root.find(".//{*}metadata")
if metadata is None:
fail(f"{archive.name} has no package metadata.")
values = {
child.tag.rsplit("}", 1)[-1]: (child.text or "").strip()
for child in metadata
if len(child) == 0
}
dependencies = {
item.get("id"): item.get("version")
for item in metadata.findall(".//{*}dependency")
}
repository = metadata.find("./{*}repository")
repository_attributes = {} if repository is None else dict(repository.attrib)
return values, dependencies, repository_attributes
def verify_checksum_file(release_dir: pathlib.Path, excluded=()) -> None:
checksum_path = release_dir / "checksums.sha256"
lines = checksum_path.read_text(encoding="utf-8").splitlines()
if not lines:
fail("checksums.sha256 is empty.")
referenced = set()
for line in lines:
digest, marker, relative = line.partition(" ")
if marker != " " or not re.fullmatch(r"[0-9a-f]{64}", digest):
fail(f"Malformed checksum line: {line}")
target = release_dir / relative
if not target.is_file():
fail(f"Checksum references missing artifact: {relative}")
actual = hashlib.sha256(target.read_bytes()).hexdigest()
if actual != digest:
fail(f"Checksum mismatch for {relative}.")
referenced.add(relative)
expected = {
path.name
for path in release_dir.iterdir()
if path.is_file()
and path.name != "checksums.sha256"
and path.name not in excluded
}
if referenced != expected:
fail(
"Checksum manifest coverage differs. "
f"Missing={expected - referenced}; extra={referenced - expected}"
)
def command_verify(args) -> None:
release_dir = pathlib.Path(args.release_dir).resolve()
version = args.version
expected = {
f"FinalFactory.Rendezvous.Client.{version}.nupkg",
f"FinalFactory.Rendezvous.Client.{version}.snupkg",
f"FinalFactory.Rendezvous.Contracts.{version}.nupkg",
f"FinalFactory.Rendezvous.Contracts.{version}.snupkg",
f"FinalFactory.Rendezvous.Server.{version}.linux-x64.tar.gz",
f"FinalFactory.Rendezvous.{version}.spdx.json",
"CHANGELOG.md",
"checksums.sha256",
"release-provenance.json",
}
checksum_exclusions = set()
if args.phase in {"publish-ready", "signing-ready", "published"}:
expected.add(f"FinalFactory.Rendezvous.Container.{version}.spdx.json")
if args.phase in {"signing-ready", "published"}:
expected.update({"container-digest.txt", "cosign.pub"})
if args.phase == "published":
expected.add("checksums.sha256.bundle")
checksum_exclusions.add("checksums.sha256.bundle")
actual = {path.name for path in release_dir.iterdir() if path.is_file()}
if actual != expected:
fail(f"Release artifact set differs. Missing={expected - actual}; extra={actual - expected}")
if args.phase in {"publish-ready", "signing-ready", "published"}:
container_sbom = load_json(
release_dir / f"FinalFactory.Rendezvous.Container.{version}.spdx.json"
)
if container_sbom.get("spdxVersion") != "SPDX-2.3":
fail("Container inventory must be an SPDX 2.3 document.")
if not container_sbom.get("packages"):
fail("Container SPDX inventory contains no packages.")
policy = load_json(pathlib.Path(args.root) / "eng" / "release-policy.json")
forbidden = tuple(fragment.lower() for fragment in policy["forbiddenArtifactNameFragments"])
for artifact in actual:
if artifact != "release-provenance.json" and any(fragment in artifact.lower() for fragment in forbidden):
fail(f"Forbidden secret-like artifact name: {artifact}")
release_sbom = load_json(
release_dir / f"FinalFactory.Rendezvous.{version}.spdx.json"
)
package_ids = {
package.get("name"): package.get("SPDXID")
for package in release_sbom.get("packages", [])
}
relationships = {
(
relationship.get("spdxElementId"),
relationship.get("relationshipType"),
relationship.get("relatedSpdxElement"),
)
for relationship in release_sbom.get("relationships", [])
}
expected_component_edges = {
("SPDXRef-Server", "SPDXRef-Contracts"),
("SPDXRef-Server", package_ids.get("LiteNetLib")),
("SPDXRef-Server", package_ids.get("Microsoft.AspNetCore.OpenApi")),
("SPDXRef-Client", "SPDXRef-Contracts"),
("SPDXRef-Client", package_ids.get("LiteNetLib")),
("SPDXRef-Contracts", package_ids.get("System.Text.Json")),
}
for source, target in expected_component_edges:
if not target or (source, "DEPENDS_ON", target) not in relationships:
fail(f"Release SPDX inventory is missing component edge {source} -> {target}.")
bcl_id = package_ids.get("Microsoft.Bcl.AsyncInterfaces")
if ("SPDXRef-Server", "DEPENDS_ON", bcl_id) in relationships:
fail("Release SPDX inventory incorrectly flattens transitive dependencies onto Server.")
for package_id in policy["publishedPackages"]:
package = release_dir / f"{package_id}.{version}.nupkg"
metadata, dependencies, repository = nuspec_metadata(package)
if metadata.get("id") != package_id or metadata.get("version") != version:
fail(f"{package.name} identity/version metadata is incorrect.")
if metadata.get("projectUrl") != "https://git.finalfactory.de/HeiKyu/Rendezvous":
fail(f"{package.name} has an incorrect project URL.")
if repository.get("url") != "https://git.finalfactory.de/HeiKyu/Rendezvous":
fail(f"{package.name} has an incorrect repository URL.")
if repository.get("commit") != provenance_commit(release_dir):
fail(f"{package.name} does not identify the release commit.")
symbol_package = release_dir / f"{package_id}.{version}.snupkg"
with zipfile.ZipFile(symbol_package) as symbols:
if not any(name.endswith(".pdb") for name in symbols.namelist()):
fail(f"{symbol_package.name} contains no portable PDB.")
with zipfile.ZipFile(package) as archive:
names = set(archive.namelist())
required_entries = {
"README.md",
"CHANGELOG.md",
f"lib/netstandard2.1/{package_id}.dll",
}
if not required_entries <= names:
fail(
f"{package.name} is missing required package content: "
f"{required_entries - names}"
)
forbidden_entries = [
name
for name in names
if any(
fragment in name.lower()
for fragment in (
"appsettings",
"launchsettings",
".env",
"credential",
"signing-key",
"private-key",
)
)
]
if forbidden_entries:
fail(
f"{package.name} contains deployable configuration or secrets: "
f"{forbidden_entries}"
)
if package_id.endswith(".Client"):
if dependencies.get("LiteNetLib") != "[2.1.4]":
fail(f"Client package must pin LiteNetLib exactly to 2.1.4: {dependencies}")
contracts_range = dependencies.get("FinalFactory.Rendezvous.Contracts", "")
if contracts_range != f"[{version}]":
fail(f"Client package does not depend on the matching Contracts version.")
provenance = load_json(release_dir / "release-provenance.json")
if provenance.get("version") != version:
fail("Release provenance does not identify the requested version.")
if provenance.get("treeState") != "clean" and not args.allow_dirty:
fail("Release provenance must identify a clean tree.")
container = provenance.get("containerImage", "")
if container.endswith(":latest") or ":latest@" in container or f":{version}" not in container:
fail(f"Container reference is mutable or not versioned: {container}")
if provenance.get("containerPlatform") != "linux/amd64":
fail("Release provenance must pin the linux/amd64 container platform.")
for field in ("containerBaseDigests", "releaseBuilderBaseDigests"):
images = provenance.get(field, [])
if not images or any("@sha256:" not in image or image.endswith(":latest") for image in images):
fail(f"Release provenance contains an unpinned build image in {field}: {images}")
build_tools = provenance.get("buildTools", [])
required_tool_prefixes = ("dotnet", "python=", "tar=", "gzip=", "jq=")
observed_tools = [f"dotnet={provenance.get('dotnetSdk', '')}", *build_tools]
for prefix in required_tool_prefixes:
if not any(tool.startswith(prefix) for tool in observed_tools):
fail(f"Release provenance is missing an artifact tool version: {prefix}")
if args.phase in {"publish-ready", "signing-ready", "published"}:
if not re.fullmatch(
r"sha256:[0-9a-f]{64}", provenance.get("containerImageId", "")
):
fail("Release provenance is missing the verified local container image ID.")
expected_namespace = (
"https://git.finalfactory.de/HeiKyu/Rendezvous/container-sbom/"
f"{version}/{provenance_commit(release_dir)}"
)
if container_sbom.get("documentNamespace") != expected_namespace:
fail("Container SPDX inventory does not identify the release commit.")
expected_created = dt.datetime.fromtimestamp(
int(provenance.get("sourceDateEpoch", 0)), dt.timezone.utc
).strftime("%Y-%m-%dT%H:%M:%SZ")
if container_sbom.get("creationInfo", {}).get("created") != expected_created:
fail("Container SPDX timestamp is not normalized to the source epoch.")
if args.phase in {"signing-ready", "published"}:
digest = (release_dir / "container-digest.txt").read_text(
encoding="utf-8"
).strip()
if not re.fullmatch(
r"git\.finalfactory\.de/heikyu/rendezvous@sha256:[0-9a-f]{64}", digest
):
fail(f"Published container digest is invalid: {digest}")
if provenance.get("containerDigest") != digest:
fail("Published provenance and container digest file differ.")
for prefix in ("docker-buildx=", "buildkit="):
if not any(tool.startswith(prefix) for tool in build_tools):
fail(f"Published provenance is missing container tool version: {prefix}")
verify_checksum_file(release_dir, checksum_exclusions)
print(f"Verified {args.phase} release artifact set for {version}.")
def provenance_commit(release_dir: pathlib.Path) -> str:
provenance = load_json(release_dir / "release-provenance.json")
commit = provenance.get("commit", "")
if not re.fullmatch(r"[0-9a-f]{40}", commit):
fail("Release provenance must contain a full Git commit SHA.")
return commit
def command_consumer(args) -> None:
assets = load_json(pathlib.Path(args.assets))
libraries = assets.get("libraries", {})
required = {
f"FinalFactory.Rendezvous.Client/{args.version}",
f"FinalFactory.Rendezvous.Contracts/{args.version}",
"LiteNetLib/2.1.4",
}
missing = required - set(libraries)
if missing:
fail(f"Consumer restore is missing exact release dependencies: {missing}")
forbidden = [name for name in libraries if name.lower().startswith("litenetlib/1.")]
if forbidden:
fail(f"Consumer resolved forbidden LiteNetLib 1.x assets: {forbidden}")
def command_consumer_config(args) -> None:
local_source = escape(str(pathlib.Path(args.local_source).resolve()))
configuration = f'''<?xml version="1.0" encoding="utf-8"?>
<configuration>
<packageSources>
<clear />
<add key="rendezvous-candidate" value="{local_source}" />
<add key="nuget.org" value="https://api.nuget.org/v3/index.json" protocolVersion="3" />
</packageSources>
<packageSourceMapping>
<packageSource key="rendezvous-candidate">
<package pattern="FinalFactory.Rendezvous.*" />
</packageSource>
<packageSource key="nuget.org">
<package pattern="*" />
</packageSource>
</packageSourceMapping>
</configuration>
'''
pathlib.Path(args.output).write_text(configuration, encoding="utf-8")
def command_source_link(args) -> None:
document = load_json(pathlib.Path(args.file))
mappings = document.get("documents", {})
expected = (
"https://git.finalfactory.de/HeiKyu/Rendezvous/raw/commit/"
f"{args.commit}/"
)
if not mappings or any(not value.startswith(expected) for value in mappings.values()):
fail(f"SourceLink mappings do not identify commit {args.commit}: {mappings}")
def command_normalize_package(args) -> None:
package_path = pathlib.Path(args.package).resolve()
if package_path.suffix not in {".nupkg", ".snupkg"}:
fail(f"Unsupported NuGet archive extension: {package_path.name}")
timestamp = dt.datetime.fromtimestamp(
int(args.source_date_epoch), dt.timezone.utc
)
zip_timestamp = (
max(timestamp.year, 1980),
timestamp.month,
timestamp.day,
timestamp.hour,
timestamp.minute,
timestamp.second - (timestamp.second % 2),
)
with zipfile.ZipFile(package_path) as source:
entries = {name: source.read(name) for name in source.namelist()}
if ".signature.p7s" in entries:
fail(f"Refusing to rewrite signed package: {package_path.name}")
core_paths = [
name
for name in entries
if name.startswith("package/services/metadata/core-properties/")
and name.endswith(".psmdcp")
]
if len(core_paths) != 1:
fail(f"Expected one NuGet core-properties part in {package_path.name}.")
entries[CORE_PROPERTIES_PATH] = entries.pop(core_paths[0])
nuspec_paths = [name for name in entries if name.endswith(".nuspec")]
if len(nuspec_paths) != 1:
fail(f"Expected one nuspec in {package_path.name}.")
nuspec = ET.fromstring(entries[nuspec_paths[0]])
nuspec_namespace = nuspec.tag.partition("}")[0].removeprefix("{")
if nuspec_namespace:
ET.register_namespace("", nuspec_namespace)
package_id = nuspec.findtext(".//{*}id")
if package_id == "FinalFactory.Rendezvous.Client":
contracts = nuspec.find(
".//{*}dependency[@id='FinalFactory.Rendezvous.Contracts']"
)
if contracts is None:
fail("Client package has no Contracts dependency to pin.")
contracts.set("version", f"[{args.version}]")
entries[nuspec_paths[0]] = ET.tostring(
nuspec, encoding="utf-8", xml_declaration=True
)
relationships_namespace = (
"http://schemas.openxmlformats.org/package/2006/relationships"
)
ET.register_namespace("", relationships_namespace)
relationships = ET.fromstring(entries["_rels/.rels"])
for relationship in relationships:
relationship_type = relationship.get("Type", "")
if relationship_type.endswith("/manifest"):
relationship.set("Id", "RManifest")
elif relationship_type.endswith("/metadata/core-properties"):
relationship.set("Id", "RCoreProperties")
relationship.set("Target", f"/{CORE_PROPERTIES_PATH}")
entries["_rels/.rels"] = ET.tostring(
relationships, encoding="utf-8", xml_declaration=True
)
temporary = package_path.with_suffix(package_path.suffix + ".normalized")
with zipfile.ZipFile(temporary, "w", compression=zipfile.ZIP_STORED) as target:
for name in sorted(entries):
info = zipfile.ZipInfo(name, date_time=zip_timestamp)
info.compress_type = zipfile.ZIP_STORED
info.create_system = 3
info.external_attr = 0o100644 << 16
target.writestr(info, entries[name])
temporary.replace(package_path)
def command_provenance(args) -> None:
versions = ET.parse(pathlib.Path(args.root) / "eng" / "Versions.props")
def version_property(name: str) -> str:
element = versions.find(f".//{name}")
if element is None or not element.text:
fail(f"Missing central release property: {name}")
return element.text.strip()
provenance = {
"schemaVersion": 1,
"version": args.version,
"commit": args.commit,
"treeState": args.tree_state,
"buildConfiguration": "Release",
"sourceDateEpoch": int(args.source_date_epoch),
"dotnetSdk": args.dotnet_sdk,
"buildTools": sorted(args.build_tool),
"containerImage": f"git.finalfactory.de/heikyu/rendezvous:{args.version}",
"containerPlatform": "linux/amd64",
"containerBaseDigests": args.base_digest,
"releaseBuilderBaseDigests": args.builder_base,
"packages": [
f"FinalFactory.Rendezvous.Client/{args.version}",
f"FinalFactory.Rendezvous.Contracts/{args.version}",
],
"compatibility": {
"minimumClientVersion": version_property("MinimumClientVersion"),
"maximumClientMajorVersion": int(version_property("MaximumClientMajorVersion")),
"httpContractVersions": [int(version_property("HttpContractVersion"))],
"udpContractVersions": [int(version_property("UdpContractVersion"))],
"connectionTicketFormatVersions": [
int(version_property("ConnectionTicketFormatVersion"))
],
"liteNetLib": version_property("LiteNetLibVersion"),
"gameplayProtocol": "exact-per-tenant",
},
}
pathlib.Path(args.output).write_text(
json.dumps(provenance, indent=2) + "\n", encoding="utf-8"
)
def command_record_container_build(args) -> None:
path = pathlib.Path(args.provenance)
provenance = load_json(path)
tools = set(provenance.get("buildTools", []))
tools.add(f"docker-buildx={args.buildx_version}")
tools.add(f"buildkit={args.buildkit_version}")
provenance["buildTools"] = sorted(tools)
provenance["containerPlatform"] = "linux/amd64"
if not re.fullmatch(r"sha256:[0-9a-f]{64}", args.image_id):
fail(f"Container image ID is invalid: {args.image_id}")
provenance["containerImageId"] = args.image_id
path.write_text(json.dumps(provenance, indent=2) + "\n", encoding="utf-8")
def command_record_container_digest(args) -> None:
if not re.fullmatch(
r"git\.finalfactory\.de/heikyu/rendezvous@sha256:[0-9a-f]{64}",
args.digest,
):
fail(f"Published container digest is invalid: {args.digest}")
release_dir = pathlib.Path(args.release_dir)
provenance_path = release_dir / "release-provenance.json"
provenance = load_json(provenance_path)
provenance["containerDigest"] = args.digest
provenance_path.write_text(
json.dumps(provenance, indent=2) + "\n", encoding="utf-8"
)
(release_dir / "container-digest.txt").write_text(
args.digest + "\n", encoding="utf-8"
)
def command_normalize_container_sbom(args) -> None:
path = pathlib.Path(args.file)
document = load_json(path)
created = dt.datetime.fromtimestamp(
int(args.source_date_epoch), dt.timezone.utc
).strftime("%Y-%m-%dT%H:%M:%SZ")
document["name"] = f"FinalFactory.Rendezvous.Container-{args.version}"
document["documentNamespace"] = (
"https://git.finalfactory.de/HeiKyu/Rendezvous/container-sbom/"
f"{args.version}/{args.commit}"
)
creation = document.setdefault("creationInfo", {})
creation["created"] = created
creation["creators"] = ["Tool: Trivy-0.69.3"]
if isinstance(document.get("packages"), list):
document["packages"] = sorted(
document["packages"],
key=lambda item: (
item.get("SPDXID", ""),
item.get("name", ""),
item.get("versionInfo", ""),
),
)
if isinstance(document.get("relationships"), list):
document["relationships"] = sorted(
document["relationships"],
key=lambda item: (
item.get("spdxElementId", ""),
item.get("relationshipType", ""),
item.get("relatedSpdxElement", ""),
),
)
path.write_text(
json.dumps(document, indent=2, sort_keys=True) + "\n", encoding="utf-8"
)
def main() -> None:
parser = argparse.ArgumentParser()
subparsers = parser.add_subparsers(dest="command", required=True)
policy = subparsers.add_parser("policy")
policy.add_argument("--root", required=True)
policy.set_defaults(handler=command_policy)
audit = subparsers.add_parser("audit")
audit.add_argument("--input", required=True)
audit.set_defaults(handler=command_audit)
sbom = subparsers.add_parser("sbom")
sbom.add_argument("--root", required=True)
sbom.add_argument("--version", required=True)
sbom.add_argument("--commit", required=True)
sbom.add_argument("--source-date-epoch", required=True)
sbom.add_argument("--server-deps", required=True)
sbom.add_argument("--output", required=True)
sbom.set_defaults(handler=command_sbom)
verify = subparsers.add_parser("verify")
verify.add_argument("--root", required=True)
verify.add_argument("--release-dir", required=True)
verify.add_argument("--version", required=True)
verify.add_argument("--allow-dirty", action="store_true")
verify.add_argument(
"--phase",
choices=("build", "publish-ready", "signing-ready", "published"),
default="build",
)
verify.set_defaults(handler=command_verify)
consumer = subparsers.add_parser("consumer")
consumer.add_argument("--assets", required=True)
consumer.add_argument("--version", required=True)
consumer.set_defaults(handler=command_consumer)
consumer_config = subparsers.add_parser("consumer-config")
consumer_config.add_argument("--local-source", required=True)
consumer_config.add_argument("--output", required=True)
consumer_config.set_defaults(handler=command_consumer_config)
source_link = subparsers.add_parser("source-link")
source_link.add_argument("--file", required=True)
source_link.add_argument("--commit", required=True)
source_link.set_defaults(handler=command_source_link)
normalize = subparsers.add_parser("normalize-package")
normalize.add_argument("--package", required=True)
normalize.add_argument("--source-date-epoch", required=True)
normalize.add_argument("--version", required=True)
normalize.set_defaults(handler=command_normalize_package)
provenance = subparsers.add_parser("provenance")
provenance.add_argument("--root", required=True)
provenance.add_argument("--version", required=True)
provenance.add_argument("--commit", required=True)
provenance.add_argument("--tree-state", required=True)
provenance.add_argument("--source-date-epoch", required=True)
provenance.add_argument("--dotnet-sdk", required=True)
provenance.add_argument("--build-tool", action="append", default=[])
provenance.add_argument("--base-digest", action="append", default=[])
provenance.add_argument("--builder-base", action="append", default=[])
provenance.add_argument("--output", required=True)
provenance.set_defaults(handler=command_provenance)
record_container = subparsers.add_parser("record-container-build")
record_container.add_argument("--provenance", required=True)
record_container.add_argument("--buildx-version", required=True)
record_container.add_argument("--buildkit-version", required=True)
record_container.add_argument("--image-id", required=True)
record_container.set_defaults(handler=command_record_container_build)
record_digest = subparsers.add_parser("record-container-digest")
record_digest.add_argument("--release-dir", required=True)
record_digest.add_argument("--digest", required=True)
record_digest.set_defaults(handler=command_record_container_digest)
normalize_container_sbom = subparsers.add_parser("normalize-container-sbom")
normalize_container_sbom.add_argument("--file", required=True)
normalize_container_sbom.add_argument("--version", required=True)
normalize_container_sbom.add_argument("--commit", required=True)
normalize_container_sbom.add_argument("--source-date-epoch", required=True)
normalize_container_sbom.set_defaults(handler=command_normalize_container_sbom)
args = parser.parse_args()
args.handler(args)
if __name__ == "__main__":
main()
+253
View File
@@ -0,0 +1,253 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:-}"
output="${2:-}"
property() {
sed -n "s:.*<$1>\(.*\)</$1>.*:\1:p" "$root/eng/Versions.props"
}
if [[ -z "$version" ]]; then
version="$(property RendezvousVersion)"
fi
semver='^(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)(-([0-9A-Za-z-]+\.)*[0-9A-Za-z-]+)?(\+([0-9A-Za-z-]+\.)*[0-9A-Za-z-]+)?$'
if [[ ! "$version" =~ $semver ]]; then
echo "Release version is not valid SemVer: $version" >&2
exit 1
fi
prerelease="${version%%+*}"
if [[ "$prerelease" == *-* ]]; then
prerelease="${prerelease#*-}"
IFS='.' read -r -a prerelease_identifiers <<<"$prerelease"
for identifier in "${prerelease_identifiers[@]}"; do
if [[ "$identifier" =~ ^[0-9]+$ && ! "$identifier" =~ ^(0|[1-9][0-9]*)$ ]]; then
echo "Numeric prerelease identifiers must not contain leading zeroes: $version" >&2
exit 1
fi
done
fi
if [[ "$version" != "$(property RendezvousVersion)" ]]; then
echo "Requested version $version differs from eng/Versions.props." >&2
exit 1
fi
for command in dotnet git python3 tar gzip sha256sum cmp jq; do
command -v "$command" >/dev/null || {
echo "Required release command is unavailable: $command" >&2
exit 1
}
done
commit="$(git -C "$root" rev-parse HEAD)"
source_date_epoch="$(git -C "$root" show -s --format=%ct "$commit")"
tree_state=clean
if [[ -n "$(git -C "$root" status --porcelain --untracked-files=normal)" ]]; then
tree_state=dirty
fi
allow_dirty=()
if [[ "$tree_state" != clean ]]; then
if [[ "${RENDEZVOUS_RELEASE_ALLOW_DIRTY:-0}" != 1 ]]; then
echo "A formal release must be built from a clean Git tree." >&2
exit 1
fi
allow_dirty=(--allow-dirty)
fi
if [[ -z "$output" ]]; then
output="$root/artifacts/release/$version"
fi
if [[ -e "$output" ]]; then
echo "Release output already exists; refusing to overwrite: $output" >&2
exit 1
fi
mkdir -p "$(dirname "$output")"
output="$(cd "$(dirname "$output")" && pwd)/$(basename "$output")"
work="$(mktemp -d "${TMPDIR:-/tmp}/rendezvous-release.XXXXXX")"
cleanup() {
rm -rf "$work"
}
trap cleanup EXIT
mkdir -p "$work/pack-1" "$work/pack-2" "$work/publish-1" "$work/publish-2" "$output"
create_server_archive() {
local publish_directory="$1"
local destination="$2"
tar --sort=name \
--mtime="@$source_date_epoch" \
--owner=0 --group=0 --numeric-owner \
-C "$publish_directory" -cf - . \
| gzip -n >"$destination"
}
common=(
-p:ContinuousIntegrationBuild=true
-p:PackageVersion="$version"
-p:RepositoryCommit="$commit"
-p:SourceRevisionId="$commit"
)
cd "$root"
dotnet restore Rendezvous.slnx --locked-mode
python3 eng/release_artifacts.py policy --root "$root" >"$work/dependency-policy.json"
dotnet package list \
--project Rendezvous.slnx \
--vulnerable \
--include-transitive \
--no-restore \
--format json >"$work/nuget-vulnerabilities.json"
python3 eng/release_artifacts.py audit --input "$work/nuget-vulnerabilities.json"
dotnet format Rendezvous.slnx --verify-no-changes --no-restore
api_before="$(sha256sum docs/api/*.json)"
dotnet build Rendezvous.slnx --configuration Release --no-restore "${common[@]}"
cp src/FinalFactory.Rendezvous.Server/bin/Release/net10.0/FinalFactory.Rendezvous.Server.dll \
"$work/FinalFactory.Rendezvous.Server.first.dll"
cp src/FinalFactory.Rendezvous.Server/bin/Release/net10.0/FinalFactory.Rendezvous.Server.pdb \
"$work/FinalFactory.Rendezvous.Server.first.pdb"
dotnet publish src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj \
--configuration Release \
--no-build \
--no-restore \
--output "$work/publish-1" \
-p:UseAppHost=false \
-p:OpenApiGenerateDocuments=false \
"${common[@]}"
create_server_archive "$work/publish-1" "$work/FinalFactory.Rendezvous.Server.first.tar.gz"
if [[ "$tree_state" == clean ]]; then
git diff --exit-code -- docs/api
elif [[ "$api_before" != "$(sha256sum docs/api/*.json)" ]]; then
echo "Generated OpenAPI changed during the release build." >&2
exit 1
fi
dotnet test Rendezvous.slnx --configuration Release --no-build
for project in Client Contracts; do
dotnet pack "src/FinalFactory.Rendezvous.$project/FinalFactory.Rendezvous.$project.csproj" \
--configuration Release --no-build --output "$work/pack-1" "${common[@]}"
done
for package in "$work/pack-1"/*; do
python3 eng/release_artifacts.py normalize-package \
--package "$package" \
--source-date-epoch "$source_date_epoch" \
--version "$version"
done
# Rebuild from the locked graph and prove package byte reproducibility.
dotnet clean Rendezvous.slnx --configuration Release >/dev/null
dotnet restore Rendezvous.slnx --locked-mode
dotnet build Rendezvous.slnx --configuration Release --no-restore "${common[@]}"
for project in Client Contracts; do
python3 eng/release_artifacts.py source-link \
--file "src/FinalFactory.Rendezvous.$project/obj/Release/netstandard2.1/FinalFactory.Rendezvous.$project.sourcelink.json" \
--commit "$commit"
done
cmp --silent \
"$work/FinalFactory.Rendezvous.Server.first.dll" \
src/FinalFactory.Rendezvous.Server/bin/Release/net10.0/FinalFactory.Rendezvous.Server.dll || {
echo "Server assembly is not byte reproducible." >&2
exit 1
}
cmp --silent \
"$work/FinalFactory.Rendezvous.Server.first.pdb" \
src/FinalFactory.Rendezvous.Server/bin/Release/net10.0/FinalFactory.Rendezvous.Server.pdb || {
echo "Server portable PDB is not byte reproducible." >&2
exit 1
}
for project in Client Contracts; do
dotnet pack "src/FinalFactory.Rendezvous.$project/FinalFactory.Rendezvous.$project.csproj" \
--configuration Release --no-build --output "$work/pack-2" "${common[@]}"
done
for package in "$work/pack-2"/*; do
python3 eng/release_artifacts.py normalize-package \
--package "$package" \
--source-date-epoch "$source_date_epoch" \
--version "$version"
done
for package in "$work/pack-1"/*; do
cmp --silent "$package" "$work/pack-2/$(basename "$package")" || {
echo "Package is not byte reproducible: $(basename "$package")" >&2
exit 1
}
done
cp "$work/pack-1"/* "$output/"
python3 eng/release_artifacts.py consumer-config \
--local-source "$output" \
--output "$work/consumer.NuGet.config"
for consumer in spacegame unscouted; do
project="$root/tests/consumers/$consumer/$(find "$root/tests/consumers/$consumer" -maxdepth 1 -name '*.csproj' -printf '%f\n')"
packages="$work/consumer-packages-$consumer"
dotnet restore "$project" \
-p:RendezvousPackageVersion="$version" \
-p:RestoreLockedMode=false \
--packages "$packages" \
--configfile "$work/consumer.NuGet.config" \
--force-evaluate
dotnet build "$project" \
--configuration Release --no-restore \
-p:RendezvousPackageVersion="$version"
assets="$(dirname "$project")/obj/project.assets.json"
python3 eng/release_artifacts.py consumer --assets "$assets" --version "$version"
done
dotnet publish src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj \
--configuration Release \
--no-build \
--no-restore \
--output "$work/publish-2" \
-p:UseAppHost=false \
-p:OpenApiGenerateDocuments=false \
"${common[@]}"
server_archive="$output/FinalFactory.Rendezvous.Server.$version.linux-x64.tar.gz"
create_server_archive "$work/publish-2" "$server_archive"
cmp --silent "$work/FinalFactory.Rendezvous.Server.first.tar.gz" "$server_archive" || {
echo "Server archive is not byte reproducible." >&2
exit 1
}
cp CHANGELOG.md "$output/CHANGELOG.md"
python3 eng/release_artifacts.py sbom \
--root "$root" \
--version "$version" \
--commit "$commit" \
--source-date-epoch "$source_date_epoch" \
--server-deps "$work/publish-2/FinalFactory.Rendezvous.Server.deps.json" \
--output "$output/FinalFactory.Rendezvous.$version.spdx.json"
provenance=(
provenance
--root "$root"
--version "$version"
--commit "$commit"
--tree-state "$tree_state"
--source-date-epoch "$source_date_epoch"
--dotnet-sdk "$(dotnet --version)"
--build-tool "python=$(python3 --version 2>&1)"
--build-tool "tar=$(tar --version | sed -n '1p')"
--build-tool "gzip=$(gzip --version | sed -n '1p')"
--build-tool "jq=$(jq --version)"
--output "$output/release-provenance.json"
)
while IFS= read -r base; do
provenance+=(--base-digest "$base")
done < <(sed -n 's/^FROM \([^ ]*\).*/\1/p' Dockerfile)
while IFS= read -r base; do
provenance+=(--builder-base "$base")
done < <(sed -n 's/^FROM \([^ ]*\).*/\1/p' eng/release-builder.Dockerfile)
python3 eng/release_artifacts.py "${provenance[@]}"
(
cd "$output"
find . -maxdepth 1 -type f ! -name checksums.sha256 -printf '%f\n' \
| LC_ALL=C sort \
| xargs sha256sum >checksums.sha256
)
python3 eng/release_artifacts.py verify \
--root "$root" \
--release-dir "$output" \
--version "$version" \
"${allow_dirty[@]}"
echo "Release artifacts verified at $output"
+56
View File
@@ -0,0 +1,56 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
base="${1:-origin/main}"
if ! git -C "$root" cat-file -e "$base^{commit}" 2>/dev/null; then
echo "Compatibility base does not exist; this is valid only for the initial version baseline: $base"
exit 0
fi
if [[ "$(git -C "$root" rev-parse "$base")" == "$(git -C "$root" rev-parse HEAD)" ]]; then
base="HEAD^"
fi
if ! git -C "$root" cat-file -e "$base:eng/Versions.props" 2>/dev/null; then
echo "Base has no release version manifest; accepting the initial compatibility baseline."
exit 0
fi
current_property() {
sed -n "s:.*<$1>\(.*\)</$1>.*:\1:p" "$root/eng/Versions.props"
}
base_property() {
git -C "$root" show "$base:eng/Versions.props" \
| sed -n "s:.*<$1>\(.*\)</$1>.*:\1:p"
}
changed() {
git -C "$root" diff --name-only "$base"...HEAD -- "$@" | grep -q .
}
require_increase() {
local property="$1"
local description="$2"
shift 2
if changed "$@"; then
local before after
before="$(base_property "$property")"
after="$(current_property "$property")"
if [[ ! "$before" =~ ^[0-9]+$ || ! "$after" =~ ^[0-9]+$ || "$after" -le "$before" ]]; then
echo "$description changed without increasing $property ($before -> $after)." >&2
exit 1
fi
fi
}
require_increase RendezvousMajorVersion "Published .NET API snapshot" \
'tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/client-public-api.txt' \
'tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/contracts-public-api.txt'
require_increase HttpContractVersion "HTTP/OpenAPI contract evidence" \
'docs/api/*.json' \
'tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/*.json' \
':(exclude)tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/connection-ticket.json'
require_increase UdpContractVersion "UDP contract evidence" \
'tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/*.hex'
require_increase ConnectionTicketFormatVersion "Connection-ticket format evidence" \
'tests/FinalFactory.Rendezvous.Tests/TestData/Contracts/v*/connection-ticket.json'
echo "Compatibility changes are paired with the required version increase."
+27
View File
@@ -0,0 +1,27 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
tag="${1:-${GITHUB_REF_NAME:-}}"
version="$(sed -n 's:.*<RendezvousVersion>\(.*\)</RendezvousVersion>.*:\1:p' "$root/eng/Versions.props")"
if [[ "$tag" != "v$version" ]]; then
echo "Release tag $tag does not match central version v$version." >&2
exit 1
fi
if [[ -n "$(git -C "$root" status --porcelain --untracked-files=normal)" ]]; then
echo "Release tag checkout is not clean." >&2
exit 1
fi
if [[ "$(git -C "$root" tag --points-at HEAD --list "$tag")" != "$tag" ]]; then
echo "Release tag $tag does not point at the checked-out commit." >&2
exit 1
fi
if ! grep -Eq "^## $version - [0-9]{4}-[0-9]{2}-[0-9]{2}$" "$root/CHANGELOG.md"; then
echo "CHANGELOG.md must contain a dated heading for $version." >&2
exit 1
fi
if grep -Eq "^## $version - Unreleased$" "$root/CHANGELOG.md"; then
echo "Release $version is still marked Unreleased." >&2
exit 1
fi
+26
View File
@@ -0,0 +1,26 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:?usage: finalize-release-candidate.sh VERSION RELEASE_DIRECTORY}"
release_dir="${2:?usage: finalize-release-candidate.sh VERSION RELEASE_DIRECTORY}"
container_sbom="$release_dir/FinalFactory.Rendezvous.Container.$version.spdx.json"
[[ -s "$container_sbom" ]] || {
echo "Container SBOM is missing or empty: $container_sbom" >&2
exit 1
}
(
cd "$release_dir"
find . -maxdepth 1 -type f ! -name checksums.sha256 -printf '%f\n' \
| LC_ALL=C sort \
| xargs sha256sum >checksums.sha256
)
python3 "$root/eng/release_artifacts.py" verify \
--root "$root" \
--release-dir "$release_dir" \
--version "$version" \
--phase publish-ready
echo "Finalized publish-ready release candidate $version"
+23
View File
@@ -0,0 +1,23 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:?usage: finalize-signing-ready-release.sh VERSION RELEASE_DIRECTORY}"
release_dir="${2:?usage: finalize-signing-ready-release.sh VERSION RELEASE_DIRECTORY}"
(
cd "$release_dir"
find . -maxdepth 1 -type f \
! -name checksums.sha256 \
! -name checksums.sha256.bundle \
-printf '%f\n' \
| LC_ALL=C sort \
| xargs sha256sum >checksums.sha256
)
python3 "$root/eng/release_artifacts.py" verify \
--root "$root" \
--release-dir "$release_dir" \
--version "$version" \
--phase signing-ready
echo "Finalized signing-ready release $version"
+77
View File
@@ -0,0 +1,77 @@
#!/usr/bin/env bash
set -euo pipefail
ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
LOCAL_KEY="${RENDEZVOUS_SMOKE_LOCAL_KEY:-$ROOT/deploy/compose/secrets/signing-key}"
if (( $# != 0 )); then
printf 'This helper accepts no arguments and mints only the fixed local Compose smoke scope.\n' >&2
exit 2
fi
command -v python3 >/dev/null || {
printf 'Missing required command: python3\n' >&2
exit 2
}
# This is deliberately a local-fixture tool, not a general credential issuer.
# Python reads the raw key from the protected file; key material never appears in
# a child process argument, environment value, temporary file, or command output.
python3 - "$LOCAL_KEY" <<'PY'
import base64
import hashlib
import hmac
import json
import os
import secrets
import stat
import sys
import time
key_path = sys.argv[1]
try:
metadata = os.lstat(key_path)
except FileNotFoundError:
raise SystemExit(f"Local Compose smoke key does not exist: {key_path}")
if stat.S_ISLNK(metadata.st_mode) or not stat.S_ISREG(metadata.st_mode):
raise SystemExit(f"Local Compose smoke key must be a regular non-symlink file: {key_path}")
parent_path = os.path.dirname(os.path.abspath(key_path))
parent = os.lstat(parent_path)
if stat.S_ISLNK(parent.st_mode) or not stat.S_ISDIR(parent.st_mode):
raise SystemExit(f"Local Compose secret directory must be a non-symlink directory: {parent_path}")
if parent.st_uid != os.geteuid() or parent.st_mode & 0o077:
raise SystemExit(f"Local Compose secret directory must be owned by this user with mode 0700: {parent_path}")
if metadata.st_uid != os.geteuid() or metadata.st_mode & 0o077 or metadata.st_nlink != 1:
raise SystemExit(f"Local Compose smoke key must be owned by this user, single-linked, and private to its owner: {key_path}")
with open(key_path, "rb") as key_file:
key = key_file.read(33)
if len(key) != 32:
raise SystemExit(f"Local Compose smoke key must be exactly 32 bytes: {key_path}")
now = int(time.time())
payload = {
"version": 1,
"issuer": "final-factory-rendezvous-smoke",
"audience": "rendezvous-service",
"subject": "local-smoke-host",
"kind": "dedicatedPublisher",
"gameId": "space-game",
"environmentId": "smoke",
"regions": ["local"],
"permissions": [],
"issuedAtUnixSeconds": now,
"notBeforeUnixSeconds": now,
"expiresAtUnixSeconds": now + 600,
"nonce": secrets.token_hex(16),
}
def base64url(value: bytes) -> str:
return base64.urlsafe_b64encode(value).rstrip(b"=").decode("ascii")
encoded = base64url(json.dumps(payload, separators=(",", ":")).encode("utf-8"))
signed = f"rv1.local-smoke-1.{encoded}"
signature = base64url(hmac.new(key, signed.encode("ascii"), hashlib.sha256).digest())
print(f"{signed}.{signature}")
PY
+153
View File
@@ -0,0 +1,153 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:?usage: publish-release.sh VERSION RELEASE_DIRECTORY}"
release_dir="${2:?usage: publish-release.sh VERSION RELEASE_DIRECTORY}"
api="${RENDEZVOUS_GITEA_API:-https://git.finalfactory.de/api/v1}"
registry="${RENDEZVOUS_CONTAINER_REGISTRY:-git.finalfactory.de}"
image="$registry/heikyu/rendezvous:$version"
token="${RENDEZVOUS_RELEASE_TOKEN:?RENDEZVOUS_RELEASE_TOKEN is required}"
username="${RENDEZVOUS_RELEASE_USERNAME:?RENDEZVOUS_RELEASE_USERNAME is required}"
release_builder="${RENDEZVOUS_RELEASE_BUILDER:?RENDEZVOUS_RELEASE_BUILDER is required}"
docker_config="$(mktemp -d)"
curl_config="$(mktemp)"
release_request=""
release_response=""
cleanup() {
[[ -z "$release_request" ]] || rm -f "$release_request"
[[ -z "$release_response" ]] || rm -f "$release_response"
rm -f "$curl_config"
rm -rf "$docker_config"
}
trap cleanup EXIT
chmod 0700 "$docker_config"
chmod 0600 "$curl_config"
printf 'header = "Authorization: token %s"\n' "$token" >"$curl_config"
export DOCKER_CONFIG="$docker_config"
for command in cosign curl docker dotnet jq; do
command -v "$command" >/dev/null || {
echo "Required publication command is unavailable: $command" >&2
exit 1
}
done
run_release_builder() {
docker run --rm \
--user "$(id -u):$(id -g)" \
--volume "$root:/source:ro" \
--volume "$release_dir:$release_dir" \
--workdir /source \
"$release_builder" "$@"
}
"$root/scripts/check-release-tag.sh" "v$version"
"$root/scripts/verify-release.sh" "$version" "$release_dir" publish-ready
require_absent() {
local description="$1"
local url="$2"
local status
status="$(curl --silent --show-error --output /dev/null --write-out '%{http_code}' \
--config "$curl_config" "$url")"
if [[ "$status" != 404 ]]; then
echo "$description must not exist before publication (HTTP $status)." >&2
exit 1
fi
}
# Gitea package versions are immutable. Require all destinations to be empty
# before the first write so a tag can never become a silent partial rerun.
require_absent "Client package $version" \
"$api/packages/HeiKyu/nuget/FinalFactory.Rendezvous.Client/$version"
require_absent "Contracts package $version" \
"$api/packages/HeiKyu/nuget/FinalFactory.Rendezvous.Contracts/$version"
require_absent "Container $version" \
"$api/packages/HeiKyu/container/rendezvous/$version"
require_absent "Release v$version" \
"$api/repos/HeiKyu/Rendezvous/releases/tags/v$version"
feed="https://git.finalfactory.de/api/packages/HeiKyu/nuget/index.json"
for package in \
"$release_dir/FinalFactory.Rendezvous.Contracts.$version.nupkg" \
"$release_dir/FinalFactory.Rendezvous.Client.$version.nupkg"; do
dotnet nuget push "$package" \
--source "$feed" \
--api-key "$token" \
--timeout 300
done
printf '%s' "$token" | docker login "$registry" --username "$username" --password-stdin
expected_image_id="$(jq -er '.containerImageId' "$release_dir/release-provenance.json")"
current_image_id="$(docker image inspect --format '{{.Id}}' "$image")"
if [[ "$current_image_id" != "$expected_image_id" ]]; then
echo "Local release tag changed after staging ($expected_image_id -> $current_image_id)." >&2
exit 1
fi
docker push "$image"
digest_ref="$(docker inspect --format '{{index .RepoDigests 0}}' "$image")"
if [[ ! "$digest_ref" =~ ^git\.finalfactory\.de/heikyu/rendezvous@sha256:[0-9a-f]{64}$ ]]; then
echo "Registry did not return an immutable Rendezvous image digest: $digest_ref" >&2
exit 1
fi
run_release_builder python3 eng/release_artifacts.py record-container-digest \
--release-dir "$release_dir" \
--digest "$digest_ref"
cosign public-key --key env://COSIGN_PRIVATE_KEY >"$release_dir/cosign.pub"
run_release_builder ./scripts/finalize-signing-ready-release.sh \
"$version" "$release_dir"
cosign sign --yes --key env://COSIGN_PRIVATE_KEY "$digest_ref"
cosign attest --yes \
--key env://COSIGN_PRIVATE_KEY \
--type https://finalfactory.de/rendezvous/release-provenance/v1 \
--predicate "$release_dir/release-provenance.json" \
"$digest_ref"
cosign sign-blob --yes \
--key env://COSIGN_PRIVATE_KEY \
--bundle "$release_dir/checksums.sha256.bundle" \
"$release_dir/checksums.sha256"
"$root/scripts/verify-release.sh" "$version" "$release_dir" published
cosign verify --key "$release_dir/cosign.pub" "$digest_ref" >/dev/null
cosign verify-attestation \
--key "$release_dir/cosign.pub" \
--type https://finalfactory.de/rendezvous/release-provenance/v1 \
"$digest_ref" >/dev/null
cosign verify-blob \
--key "$release_dir/cosign.pub" \
--bundle "$release_dir/checksums.sha256.bundle" \
"$release_dir/checksums.sha256" >/dev/null
release_request="$(mktemp)"
release_response="$(mktemp)"
prerelease=false
if [[ "$version" == *-* ]]; then
prerelease=true
fi
jq -n \
--arg tag "v$version" \
--arg commit "${GITHUB_SHA:?GITHUB_SHA is required}" \
--arg digest "$digest_ref" \
--argjson prerelease "$prerelease" \
--rawfile changelog "$release_dir/CHANGELOG.md" \
'{tag_name:$tag,target_commitish:$commit,name:("Rendezvous " + $tag),body:($changelog + "\n\n## Immutable container\n\n`" + $digest + "`\n"),draft:false,prerelease:$prerelease}' \
>"$release_request"
curl --fail --silent --show-error \
--request POST \
--config "$curl_config" \
--header 'Content-Type: application/json' \
--data-binary "@$release_request" \
"$api/repos/HeiKyu/Rendezvous/releases" >"$release_response"
release_id="$(jq -er '.id' "$release_response")"
for artifact in "$release_dir"/*; do
curl --fail --silent --show-error \
--request POST \
--config "$curl_config" \
--form "attachment=@$artifact" \
"$api/repos/HeiKyu/Rendezvous/releases/$release_id/assets?name=$(basename "$artifact")" \
>/dev/null
done
echo "Published immutable release v$version with container $digest_ref"
+58
View File
@@ -0,0 +1,58 @@
#!/usr/bin/env bash
set -euo pipefail
ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
PROFILE="${RENDEZVOUS_CAPACITY_PROFILE:-quick}"
OUTPUT="${RENDEZVOUS_CAPACITY_OUTPUT:-$ROOT/artifacts/capacity/rendezvous-capacity-v2.json}"
CPUSET="${RENDEZVOUS_CAPACITY_CPUSET:-}"
PROJECT="$ROOT/tests/FinalFactory.Rendezvous.Capacity/FinalFactory.Rendezvous.Capacity.csproj"
TESTS="$ROOT/tests/FinalFactory.Rendezvous.Tests/FinalFactory.Rendezvous.Tests.csproj"
if [[ "$PROFILE" != quick && "$PROFILE" != candidate ]]; then
printf 'RENDEZVOUS_CAPACITY_PROFILE must be quick or candidate.\n' >&2
exit 2
fi
command -v dotnet >/dev/null || {
printf 'Missing required command: dotnet\n' >&2
exit 2
}
if [[ -n "$CPUSET" ]]; then
command -v taskset >/dev/null || {
printf 'taskset is required when RENDEZVOUS_CAPACITY_CPUSET is set.\n' >&2
exit 2
}
fi
cd "$ROOT"
export RENDEZVOUS_EVIDENCE_COMMIT="$(git rev-parse HEAD)"
if [[ -n "$(git status --porcelain)" ]]; then
export RENDEZVOUS_EVIDENCE_TREE_STATE=dirty
else
export RENDEZVOUS_EVIDENCE_TREE_STATE=clean
fi
if [[ "$PROFILE" == candidate && "$RENDEZVOUS_EVIDENCE_TREE_STATE" != clean ]]; then
printf 'Candidate evidence requires a clean source tree.\n' >&2
exit 2
fi
export RENDEZVOUS_EVIDENCE_CPUSET="${CPUSET:-unrestricted}"
export RENDEZVOUS_EVIDENCE_COMMAND="RENDEZVOUS_CAPACITY_PROFILE=$PROFILE RENDEZVOUS_CAPACITY_CPUSET=${CPUSET:-unrestricted} ./scripts/run-capacity-gate.sh"
dotnet restore "$ROOT/Rendezvous.slnx" --locked-mode
dotnet build "$ROOT/Rendezvous.slnx" --configuration Release --no-restore
filter='FullyQualifiedName~TrackerCapacityFailsClosedWithoutGrowingAndAWindowResetRecovers|FullyQualifiedName~OptionalTrafficCannotConsumeTheLeaseOperationReserve|FullyQualifiedName~ConcurrentAbusiveBurstStaysBoundedAndCannotBlockCriticalHttp|FullyQualifiedName~HttpOverloadIsTypedAndOversizedBodiesAreRejectedBeforeDispatch|FullyQualifiedName~WallClockMovementDoesNotExpireOrExtendLease|FullyQualifiedName~RepeatedMutableDeadlineRefreshesKeepOneScheduledEntryPerKey|FullyQualifiedName~RepeatedPrincipalRevocationCanExtendButCannotShortenProtection|FullyQualifiedName~RestartHasNewGenerationAndNoEphemeralState|FullyQualifiedName~RestartReturnsTypedUnavailabilityThenAllowsHostReregistration|FullyQualifiedName~DrainRejectsNewWorkAllowsInflightCompletionThenClearsState|FullyQualifiedName~UnavailableStoreFailsNewAuthorizationClosedAndErasesActiveState|FullyQualifiedName~KeyRotationHonorsOverlapAndRejectsRetiredKeys|FullyQualifiedName~OperatorSurfaceSeparatesAuthenticationConfirmsActionsAndRedactsInspection|FullyQualifiedName~SigtermDrainsThenReleasesHttpAndUdpSockets|FullyQualifiedName~ProductionTransportSoakKeepsHandlesMemoryAndSocketsBounded|FullyQualifiedName~NativeLiteNetLibRequestsIntroduceTheAuthorizedPair'
dotnet test "$TESTS" --configuration Release --no-build --filter "$filter" \
--logger 'console;verbosity=minimal'
mkdir -p "$(dirname "$OUTPUT")"
arguments=(
dotnet run --project "$PROJECT" --configuration Release --no-build --
--profile "$PROFILE" --output "$OUTPUT"
)
if [[ -n "$CPUSET" ]]; then
taskset -c "$CPUSET" "${arguments[@]}"
else
"${arguments[@]}"
fi
printf 'Capacity and resilience gate passed; evidence: %s\n' "$OUTPUT"
+123
View File
@@ -0,0 +1,123 @@
#!/usr/bin/env bash
set -euo pipefail
ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
SERVICE_URL="${RENDEZVOUS_SMOKE_HTTP_URL:-http://127.0.0.1:8080/}"
MEDIATOR="${RENDEZVOUS_SMOKE_UDP_ENDPOINT:-127.0.0.1:9050}"
TIMEOUT_SECONDS="${RENDEZVOUS_SMOKE_TIMEOUT_SECONDS:-30}"
LOCAL_KEY="${RENDEZVOUS_SMOKE_LOCAL_KEY:-$ROOT/deploy/compose/secrets/signing-key}"
PROJECT="$ROOT/src/FinalFactory.Rendezvous.TestClient/FinalFactory.Rendezvous.TestClient.csproj"
BUILD_CONFIGURATION="${RENDEZVOUS_SMOKE_CONFIGURATION:-Release}"
GAME_ID="${RENDEZVOUS_SMOKE_GAME_ID:-space-game}"
ENVIRONMENT_ID="${RENDEZVOUS_SMOKE_ENVIRONMENT_ID:-smoke}"
REGION="${RENDEZVOUS_SMOKE_REGION:-local}"
PROTOCOL_VERSION="${RENDEZVOUS_SMOKE_PROTOCOL_VERSION:-1}"
for command in curl dotnet jq mktemp tail; do
command -v "$command" >/dev/null || {
printf 'Missing required command: %s\n' "$command" >&2
exit 2
}
done
if [[ ! "$TIMEOUT_SECONDS" =~ ^[0-9]+$ ]] || (( TIMEOUT_SECONDS < 1 || TIMEOUT_SECONDS > 300 )); then
printf 'RENDEZVOUS_SMOKE_TIMEOUT_SECONDS must be an integer from 1 through 300.\n' >&2
exit 2
fi
if [[ ! "$PROTOCOL_VERSION" =~ ^[0-9]+$ ]] || (( PROTOCOL_VERSION < 1 )); then
printf 'RENDEZVOUS_SMOKE_PROTOCOL_VERSION must be a positive integer.\n' >&2
exit 2
fi
for scoped_value in "$GAME_ID" "$ENVIRONMENT_ID" "$REGION"; do
if [[ -z "$scoped_value" ]]; then
printf 'Smoke game, environment, and region values must not be empty.\n' >&2
exit 2
fi
done
local_credential() {
if [[ "$GAME_ID" != space-game || "$ENVIRONMENT_ID" != smoke \
|| "$REGION" != local || "$PROTOCOL_VERSION" != 1 ]]; then
printf 'The local credential helper supports only space-game/smoke/local protocol 1. Supply RENDEZVOUS_PUBLISHER_CREDENTIAL for any other scope.\n' >&2
exit 2
fi
RENDEZVOUS_SMOKE_LOCAL_KEY="$LOCAL_KEY" \
"$ROOT/scripts/mint-local-publisher-credential.sh"
}
credential="${RENDEZVOUS_PUBLISHER_CREDENTIAL:-}"
if [[ -z "$credential" ]]; then
credential="$(local_credential)"
fi
export RENDEZVOUS_PUBLISHER_CREDENTIAL="$credential"
curl --fail --silent --show-error --max-time 5 "${SERVICE_URL%/}/health/live" >/dev/null
curl --fail --silent --show-error --max-time 5 "${SERVICE_URL%/}/health/ready" >/dev/null
temp_dir="$(mktemp -d)"
host_log="$temp_dir/host.jsonl"
join_log="$temp_dir/join.jsonl"
host_pid=''
cleanup() {
local status="$?"
if [[ -n "$host_pid" ]] && kill -0 "$host_pid" 2>/dev/null; then
kill -TERM "$host_pid" 2>/dev/null || true
wait "$host_pid" 2>/dev/null || true
fi
if [[ "$status" -ne 0 ]]; then
printf 'Deployment smoke failed; sanitized diagnostic events follow.\n' >&2
[[ -f "$host_log" ]] && jq -c . "$host_log" >&2 || true
[[ -f "$join_log" ]] && jq -c . "$join_log" >&2 || true
fi
rm -rf "$temp_dir"
return "$status"
}
trap cleanup EXIT
trap 'exit 130' INT
trap 'exit 143' TERM
dotnet run --project "$PROJECT" --configuration "$BUILD_CONFIGURATION" --no-build -- \
host --service "$SERVICE_URL" --mediator "$MEDIATOR" \
--game "$GAME_ID" --environment "$ENVIRONMENT_ID" --region "$REGION" --protocol "$PROTOCOL_VERSION" \
--script --json --exit-after-echo --timeout-seconds "$TIMEOUT_SECONDS" \
>"$host_log" 2>&1 &
host_pid="$!"
ready=false
for ((iteration = 0; iteration < TIMEOUT_SECONDS * 4; iteration++)); do
if jq -e 'select(.event == "host.ready")' "$host_log" >/dev/null 2>&1; then
ready=true
break
fi
if ! kill -0 "$host_pid" 2>/dev/null; then
printf 'Host diagnostic stopped before it became ready.\n' >&2
jq -c . "$host_log" >&2 || true
exit 1
fi
sleep 0.25
done
if [[ "$ready" != true ]]; then
printf 'Host diagnostic did not become ready within %s seconds.\n' "$TIMEOUT_SECONDS" >&2
exit 1
fi
listing_id="$(jq -r 'select(.event == "host.registered") | .listingId' "$host_log" | tail -n 1)"
if [[ -z "$listing_id" || "$listing_id" == null ]]; then
printf 'Host diagnostic did not report a listing ID.\n' >&2
exit 1
fi
dotnet run --project "$PROJECT" --configuration "$BUILD_CONFIGURATION" --no-build -- \
join --service "$SERVICE_URL" --mediator "$MEDIATOR" \
--game "$GAME_ID" --environment "$ENVIRONMENT_ID" --region "$REGION" --protocol "$PROTOCOL_VERSION" \
--listing "$listing_id" --script --json --timeout-seconds "$TIMEOUT_SECONDS" \
>"$join_log" 2>&1
wait "$host_pid"
host_pid=''
jq -e 'select(.event == "host.direct-traffic" and .status == "verified")' "$host_log" >/dev/null
jq -e 'select(.event == "host.deregistered" and .status == "complete")' "$host_log" >/dev/null
jq -e 'select(.event == "join.direct-traffic" and .status == "verified")' "$join_log" >/dev/null
jq -e 'select(.event == "join.outcome-report" and .status == "accepted")' "$join_log" >/dev/null
printf 'Rendezvous deployment smoke passed: HTTP live/ready and authenticated UDP mediation/direct traffic.\n'
+78
View File
@@ -0,0 +1,78 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:?usage: verify-real-consumers.sh VERSION RELEASE_DIRECTORY}"
release_dir="${2:?usage: verify-real-consumers.sh VERSION RELEASE_DIRECTORY}"
manifest="$root/eng/consumer-revisions.json"
work="$(mktemp -d "${TMPDIR:-/tmp}/rendezvous-consumers.XXXXXX")"
cleanup() {
rm -rf "$work"
}
trap cleanup EXIT
for command in dotnet git jq python3; do
command -v "$command" >/dev/null || {
echo "Required consumer verification command is unavailable: $command" >&2
exit 1
}
done
python3 "$root/eng/release_artifacts.py" consumer-config \
--local-source "$release_dir" \
--output "$work/NuGet.config"
count="$(jq '.consumers | length' "$manifest")"
for ((index = 0; index < count; index++)); do
name="$(jq -r ".consumers[$index].name" "$manifest")"
repository="$(jq -r ".consumers[$index].repository" "$manifest")"
revision="$(jq -r ".consumers[$index].revision" "$manifest")"
project_relative="$(jq -r ".consumers[$index].project" "$manifest")"
checkout="$work/$name"
git -c init.defaultBranch=main init --quiet "$checkout"
git -C "$checkout" remote add origin "$repository"
git -C "$checkout" fetch --quiet --depth 1 origin "$revision"
GIT_LFS_SKIP_SMUDGE=1 git -C "$checkout" checkout --quiet --detach FETCH_HEAD
[[ "$(git -C "$checkout" rev-parse HEAD)" == "$revision" ]] || {
echo "$name did not resolve the pinned consumer revision." >&2
exit 1
}
project="$checkout/$project_relative"
[[ -f "$project" ]] || {
echo "$name consumer project does not exist at $project_relative." >&2
exit 1
}
targets="$work/$name.Rendezvous.Consumer.targets"
cat >"$targets" <<EOF
<Project>
<ItemGroup Condition="'\$(MSBuildProjectFullPath)' == '$project'">
<PackageReference Include="FinalFactory.Rendezvous.Client" Version="[$version]" />
<PackageReference Include="FinalFactory.Rendezvous.Contracts" Version="[$version]" />
</ItemGroup>
</Project>
EOF
packages="$work/packages-$name"
dotnet restore "$project" \
-p:CustomAfterMicrosoftCommonTargets="$targets" \
-p:RestorePackagesWithLockFile=false \
-p:RestoreLockedMode=false \
--packages "$packages" \
--configfile "$work/NuGet.config" \
--force-evaluate
assets=""
while IFS= read -r candidate_assets; do
if grep -Fq "FinalFactory.Rendezvous.Client/$version" "$candidate_assets"; then
assets="$candidate_assets"
break
fi
done < <(find "$checkout" -path '*/obj/project.assets.json' -type f -print)
[[ -n "$assets" ]] || {
echo "$name restore did not produce assets for the injected Rendezvous references." >&2
exit 1
}
python3 "$root/eng/release_artifacts.py" consumer \
--assets "$assets" \
--version "$version"
echo "Verified $name at $revision can pin and restore Rendezvous $version."
done
+13
View File
@@ -0,0 +1,13 @@
#!/usr/bin/env bash
set -euo pipefail
root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
version="${1:?usage: verify-release.sh VERSION [RELEASE_DIRECTORY] [PHASE]}"
release_dir="${2:-$root/artifacts/release/$version}"
phase="${3:-build}"
python3 "$root/eng/release_artifacts.py" verify \
--root "$root" \
--release-dir "$release_dir" \
--version "$version" \
--phase "$phase"
@@ -7,10 +7,12 @@
<PackageId>FinalFactory.Rendezvous.Client</PackageId>
<PackageReadmeFile>README.md</PackageReadmeFile>
<Description>Godot-independent client SDK for Final Factory Rendezvous.</Description>
<PackageTags>final-factory;multiplayer;nat;godot;litenetlib</PackageTags>
</PropertyGroup>
<ItemGroup>
<ProjectReference Include="../FinalFactory.Rendezvous.Contracts/FinalFactory.Rendezvous.Contracts.csproj" />
<PackageReference Include="LiteNetLib" />
<None Update="README.md" Pack="true" PackagePath="\" />
<None Include="../../CHANGELOG.md" Pack="true" PackagePath="\" Link="CHANGELOG.md" />
</ItemGroup>
</Project>
@@ -75,7 +75,8 @@ public sealed class RendezvousJoinClient : IRendezvousJoinClient
RendezvousConnectionOutcomeSource.Caller,
RendezvousConnectionFailureCategory.Lifecycle,
RendezvousConnectionPhase.Authorization,
elapsed.Elapsed));
elapsed.Elapsed,
dedicatedFallback));
}
}
@@ -67,12 +67,17 @@ factory does not open a socket, and synchronized events must remain enabled:
```csharp
RendezvousNetListener networkEvents = new();
NetManager gameplayNetManager = networkEvents.CreateManager();
gameplayNetManager.ChannelsCount = 3; // example: configure the game protocol first
if (!gameplayNetManager.Start(0))
{
throw new InvalidOperationException("The gameplay UDP socket could not start.");
}
```
LiteNetLib defaults to one QoS channel. Set `ChannelsCount` before `Start` when
the game protocol uses more than one; both game processes must agree. Rendezvous
does not reserve or reinterpret any gameplay channel.
The host polls join invitations asynchronously; that method only queues a
snapshot and never calls the manager. `Poll()` is the sole SDK path that invokes
LiteNetLib and dispatches its synchronized callbacks. Call it once per game
@@ -4,7 +4,7 @@
".NETStandard,Version=v2.1": {
"LiteNetLib": {
"type": "Direct",
"requested": "[2.1.4, )",
"requested": "[2.1.4, 2.1.4]",
"resolved": "2.1.4",
"contentHash": "KWlxvMw3Urpqj9joD96LRiK+LC62pQNs/zkXRJc+rHnxgkGp+vV703xzDrxRmv+V1YhCFfIGzs5nrVWtREIlyA=="
},
@@ -5,9 +5,13 @@
<RootNamespace>FinalFactory.Rendezvous.Contracts</RootNamespace>
<IsPackable>true</IsPackable>
<PackageId>FinalFactory.Rendezvous.Contracts</PackageId>
<PackageReadmeFile>README.md</PackageReadmeFile>
<Description>Versioned transport-neutral contracts for Final Factory Rendezvous.</Description>
<PackageTags>final-factory;multiplayer;contracts;godot</PackageTags>
</PropertyGroup>
<ItemGroup>
<PackageReference Include="System.Text.Json" />
<None Update="README.md" Pack="true" PackagePath="\" />
<None Include="../../CHANGELOG.md" Pack="true" PackagePath="\" Link="CHANGELOG.md" />
</ItemGroup>
</Project>
@@ -0,0 +1,8 @@
# FinalFactory.Rendezvous.Contracts
Transport-neutral v1 HTTP/UDP contract types for Final Factory Rendezvous.
The package targets `netstandard2.1`, contains no Godot or LiteNetLib dependency,
and is versioned with the Client package and server release.
Compatibility and migration policy is maintained in the repository's
`docs/releases/README.md` document.
@@ -0,0 +1,111 @@
using System.ComponentModel.DataAnnotations;
namespace FinalFactory.Rendezvous.Server.Abuse;
internal sealed class AbuseProtectionOptions
{
public const string SectionName = "Rendezvous:AbuseProtection";
[Range(1, 60)]
public int WindowSeconds { get; set; } = 1;
[Range(1_000, 1_000_000)]
public int MaxTrackedKeys { get; set; } = 100_000;
[Range(0, 100_000)]
public int CriticalTrackedKeyReserve { get; set; } = 2_048;
[Range(1_000, 999_999)]
public int UdpTrackedKeyLimit { get; set; } = 70_000;
public string[] TrustedProxyAddresses { get; set; } = [];
public string[] OperatorAllowedAddresses { get; set; } = [];
[Range(1, 100_000)]
public int HealthGlobalRequestsPerWindow { get; set; } = 1_000;
[Range(1, 10_000)]
public int HealthGlobalConcurrency { get; set; } = 32;
[Range(1, 100_000)]
public int HealthIpPrefixRequestsPerWindow { get; set; } = 120;
[Range(1, 1_000)]
public int HealthIpPrefixConcurrency { get; set; } = 8;
[Range(1, 100_000)]
public int OperatorGlobalRequestsPerWindow { get; set; } = 1_000;
[Range(1, 10_000)]
public int OperatorGlobalConcurrency { get; set; } = 32;
[Range(1, 100_000)]
public int OperatorIpPrefixRequestsPerWindow { get; set; } = 120;
[Range(1, 1_000)]
public int OperatorIpPrefixConcurrency { get; set; } = 8;
[Range(1, 1_000_000)]
public int HttpGlobalRequestsPerWindow { get; set; } = 20_000;
[Range(1, 1_000_000)]
public int HttpOptionalRequestsPerWindow { get; set; } = 18_000;
[Range(1, 100_000)]
public int HttpIpPrefixRequestsPerWindow { get; set; } = 500;
[Range(1, 100_000)]
public int HttpOptionalIpPrefixRequestsPerWindow { get; set; } = 450;
[Range(1, 1_000_000)]
public int HttpOperationRequestsPerWindow { get; set; } = 5_000;
[Range(1, 1_000_000)]
public int HttpTenantRequestsPerWindow { get; set; } = 2_000;
[Range(1, 100_000)]
public int HttpPrincipalRequestsPerWindow { get; set; } = 500;
[Range(1, 100_000)]
public int HttpResourceRequestsPerWindow { get; set; } = 200;
[Range(1, 100_000)]
public int HttpGlobalConcurrency { get; set; } = 1_024;
[Range(1, 100_000)]
public int HttpOptionalConcurrency { get; set; } = 768;
[Range(1, 10_000)]
public int HttpIpPrefixConcurrency { get; set; } = 64;
[Range(1, 10_000)]
public int HttpOptionalIpPrefixConcurrency { get; set; } = 48;
[Range(1, 100_000)]
public int HttpOperationConcurrency { get; set; } = 256;
[Range(1, 100_000)]
public int HttpTenantConcurrency { get; set; } = 256;
[Range(1, 10_000)]
public int HttpPrincipalConcurrency { get; set; } = 32;
[Range(1, 10_000)]
public int HttpResourceConcurrency { get; set; } = 16;
[Range(1, 10_000_000)]
public int UdpGlobalDatagramsPerWindow { get; set; } = 100_000;
[Range(1, 1_000_000)]
public int UdpIpPrefixDatagramsPerWindow { get; set; } = 2_000;
[Range(1, 10_000_000)]
public int UdpOperationDatagramsPerWindow { get; set; } = 50_000;
[Range(1, 100_000)]
public int UdpCapabilityDatagramsPerWindow { get; set; } = 120;
[Range(1, 100_000)]
public int UdpResourceDatagramsPerWindow { get; set; } = 240;
}
@@ -0,0 +1,530 @@
using System.Buffers;
using System.Net;
using System.Security.Cryptography;
using System.Text;
using FinalFactory.Rendezvous.Server.Observability;
using Microsoft.Extensions.Options;
namespace FinalFactory.Rendezvous.Server.Abuse;
internal sealed class AbuseProtectionService
{
private readonly AbuseProtectionOptions _options;
private readonly TimeProvider _timeProvider;
private readonly TrackerState _httpTracker;
private readonly TrackerState _udpTracker;
private readonly RendezvousTelemetry? _telemetry;
private readonly HashSet<string> _operatorAllowedAddresses;
public AbuseProtectionService(
IOptions<AbuseProtectionOptions> options,
TimeProvider? timeProvider = null,
RendezvousTelemetry? telemetry = null)
{
_options = options.Value;
_timeProvider = timeProvider ?? TimeProvider.System;
_telemetry = telemetry;
_operatorAllowedAddresses = options.Value.OperatorAllowedAddresses
.Select(static value => IPAddress.TryParse(value, out IPAddress? address)
? NormalizeAddress(address).ToString()
: string.Empty)
.Where(static value => value.Length > 0)
.ToHashSet(StringComparer.Ordinal);
DateTimeOffset now = _timeProvider.GetUtcNow();
_httpTracker = new(now);
_udpTracker = new(now);
}
public bool TryAcquireHttpIngress(
IPAddress? remoteAddress,
string operation,
out AbuseLease? lease,
out int retryAfterSeconds)
{
string prefix = GetNetworkPrefix(remoteAddress);
List<RateDimension> rates =
[
new("http:rate:global", _options.HttpGlobalRequestsPerWindow),
new($"http:rate:ip:{prefix}", _options.HttpIpPrefixRequestsPerWindow),
new($"http:rate:operation:{operation}", _options.HttpOperationRequestsPerWindow),
];
List<RateDimension> concurrency =
[
new("http:concurrency:global", _options.HttpGlobalConcurrency),
new($"http:concurrency:ip:{prefix}", _options.HttpIpPrefixConcurrency),
new($"http:concurrency:operation:{operation}", _options.HttpOperationConcurrency),
];
if (!IsLeaseCriticalOperation(operation))
{
rates.Add(new("http:rate:optional", _options.HttpOptionalRequestsPerWindow));
rates.Add(new($"http:rate:optional-ip:{prefix}",
_options.HttpOptionalIpPrefixRequestsPerWindow));
concurrency.Add(new("http:concurrency:optional", _options.HttpOptionalConcurrency));
concurrency.Add(new($"http:concurrency:optional-ip:{prefix}",
_options.HttpOptionalIpPrefixConcurrency));
}
return TryAcquire(
[.. rates],
[.. concurrency],
TrackerDomain.Http,
IsLeaseCriticalOperation(operation),
out lease,
out retryAfterSeconds);
}
public bool TryAcquireHealthIngress(
IPAddress? remoteAddress,
out AbuseLease? lease,
out int retryAfterSeconds)
{
string prefix = GetNetworkPrefix(remoteAddress);
RateDimension[] rates =
[
new("health:rate:global", _options.HealthGlobalRequestsPerWindow),
new($"health:rate:ip:{prefix}", _options.HealthIpPrefixRequestsPerWindow),
];
RateDimension[] concurrency =
[
new("health:concurrency:global", _options.HealthGlobalConcurrency),
new($"health:concurrency:ip:{prefix}", _options.HealthIpPrefixConcurrency),
];
return TryAcquire(
rates,
concurrency,
TrackerDomain.Http,
true,
out lease,
out retryAfterSeconds);
}
public bool IsOperatorSourceAllowed(IPAddress? remoteAddress) =>
remoteAddress is not null
&& _operatorAllowedAddresses.Contains(NormalizeAddress(remoteAddress).ToString());
public bool TryAcquireOperatorIngress(
IPAddress? remoteAddress,
out AbuseLease? lease,
out int retryAfterSeconds)
{
string prefix = GetNetworkPrefix(remoteAddress);
RateDimension[] rates =
[
new("operator:rate:global", _options.OperatorGlobalRequestsPerWindow),
new($"operator:rate:ip:{prefix}", _options.OperatorIpPrefixRequestsPerWindow),
];
RateDimension[] concurrency =
[
new("operator:concurrency:global", _options.OperatorGlobalConcurrency),
new($"operator:concurrency:ip:{prefix}", _options.OperatorIpPrefixConcurrency),
];
return TryAcquire(
rates,
concurrency,
TrackerDomain.Http,
true,
out lease,
out retryAfterSeconds);
}
public bool TryAcquireHttpIdentity(
string operation,
string? tenant,
string? principal,
string? resource,
out AbuseLease? lease,
out int retryAfterSeconds) => TryAcquireHttpIdentity(
operation,
null,
tenant,
principal,
resource,
out lease,
out retryAfterSeconds);
public bool TryAcquireHttpIdentity(
string operation,
IPAddress? remoteAddress,
string? tenant,
string? principal,
string? resource,
out AbuseLease? lease,
out int retryAfterSeconds)
{
string sourcePrefix = GetNetworkPrefix(remoteAddress);
List<RateDimension> rates = [];
List<RateDimension> concurrency = [];
AddDimension(rates, concurrency, "tenant", tenant,
_options.HttpTenantRequestsPerWindow, _options.HttpTenantConcurrency);
AddDimension(rates, concurrency, "principal", principal,
_options.HttpPrincipalRequestsPerWindow, _options.HttpPrincipalConcurrency);
AddDimension(rates, concurrency, "resource", resource,
_options.HttpResourceRequestsPerWindow, _options.HttpResourceConcurrency);
return TryAcquire(
[.. rates],
[.. concurrency],
TrackerDomain.Http,
IsLeaseCriticalOperation(operation),
out lease,
out retryAfterSeconds);
void AddDimension(
List<RateDimension> rateDimensions,
List<RateDimension> concurrencyDimensions,
string kind,
string? value,
int rateLimit,
int concurrencyLimit)
{
if (string.IsNullOrEmpty(value))
{
return;
}
if (kind == "resource")
{
string validationKey =
$"http:source-resource:{operation}:{sourcePrefix}:{value}";
rateDimensions.Add(new($"{validationKey}:rate", rateLimit));
concurrencyDimensions.Add(new($"{validationKey}:concurrency", concurrencyLimit));
}
string key = kind == "resource"
? $"http:resource-scoped:{operation}:{tenant ?? string.Empty}|{principal ?? string.Empty}:{value}"
: $"http:{kind}:{operation}:{value}";
rateDimensions.Add(new($"{key}:rate", rateLimit));
concurrencyDimensions.Add(new($"{key}:concurrency", concurrencyLimit));
}
}
public bool TryAcceptUdpIngress(IPAddress? remoteAddress, string operation)
{
string prefix = GetNetworkPrefix(remoteAddress);
RateDimension[] rates =
[
new("udp:rate:global", _options.UdpGlobalDatagramsPerWindow),
new($"udp:rate:ip:{prefix}", _options.UdpIpPrefixDatagramsPerWindow),
new($"udp:rate:operation:{operation}", _options.UdpOperationDatagramsPerWindow),
];
return TryAcquire(
rates,
[],
TrackerDomain.Udp,
false,
out AbuseLease? lease,
out _)
&& DisposeAccepted(lease);
}
public bool TryAcceptUdpIdentity(
string operation,
string capability,
string resource) => TryAcceptUdpIdentity(
operation,
null,
capability,
resource);
public bool TryAcceptUdpIdentity(
string operation,
IPAddress? remoteAddress,
string capability,
string resource)
{
string sourcePrefix = GetNetworkPrefix(remoteAddress);
string capabilityFingerprint = FingerprintSecret(capability);
RateDimension[] rates =
[
new($"udp:rate:capability:{operation}:{capabilityFingerprint}",
_options.UdpCapabilityDatagramsPerWindow),
new($"udp:rate:source-resource:{operation}:{sourcePrefix}:{resource}",
_options.UdpResourceDatagramsPerWindow),
new($"udp:rate:resource:{operation}:{capabilityFingerprint}:{resource}",
_options.UdpResourceDatagramsPerWindow),
];
return TryAcquire(
rates,
[],
TrackerDomain.Udp,
false,
out AbuseLease? lease,
out _)
&& DisposeAccepted(lease);
}
public static string FingerprintSecret(string secret)
{
int byteCount = Encoding.UTF8.GetByteCount(secret);
byte[]? rented = null;
Span<byte> encoded = byteCount <= 1_024
? stackalloc byte[byteCount]
: (rented = ArrayPool<byte>.Shared.Rent(byteCount)).AsSpan(0, byteCount);
Span<byte> digest = stackalloc byte[32];
try
{
_ = Encoding.UTF8.GetBytes(secret, encoded);
_ = SHA256.HashData(encoded, digest);
return Convert.ToHexString(digest[..12]);
}
finally
{
CryptographicOperations.ZeroMemory(encoded);
CryptographicOperations.ZeroMemory(digest);
if (rented is not null)
{
ArrayPool<byte>.Shared.Return(rented);
}
}
}
internal int TrackedKeyCount
{
get
{
int http;
int udp;
lock (_httpTracker.Gate)
{
http = _httpTracker.WindowCounts.Count + _httpTracker.ConcurrencyCounts.Count;
}
lock (_udpTracker.Gate)
{
udp = _udpTracker.WindowCounts.Count + _udpTracker.ConcurrencyCounts.Count;
}
return http + udp;
}
}
private bool TryAcquire(
ReadOnlySpan<RateDimension> rates,
ReadOnlySpan<RateDimension> concurrency,
TrackerDomain domain,
bool canUseCriticalReserve,
out AbuseLease? lease,
out int retryAfterSeconds)
{
TrackerState tracker = domain == TrackerDomain.Udp ? _udpTracker : _httpTracker;
bool accepted;
lock (tracker.Gate)
{
accepted = TryAcquireLocked(
tracker,
rates,
concurrency,
domain,
canUseCriticalReserve,
out lease,
out retryAfterSeconds);
}
if (!accepted)
{
_telemetry?.RecordLimiterDrop(
domain == TrackerDomain.Udp ? "udp" : "http",
"rate-or-concurrency");
}
return accepted;
}
private bool TryAcquireLocked(
TrackerState tracker,
ReadOnlySpan<RateDimension> rates,
ReadOnlySpan<RateDimension> concurrency,
TrackerDomain domain,
bool canUseCriticalReserve,
out AbuseLease? lease,
out int retryAfterSeconds)
{
DateTimeOffset now = _timeProvider.GetUtcNow();
TimeSpan window = TimeSpan.FromSeconds(_options.WindowSeconds);
if (now - tracker.WindowStartedAt >= window || now < tracker.WindowStartedAt)
{
tracker.WindowCounts.Clear();
tracker.WindowStartedAt = now;
}
retryAfterSeconds = Math.Max(
1,
(int)Math.Ceiling((window - (now - tracker.WindowStartedAt)).TotalSeconds));
int stagedNewKeys = 0;
int partitionLimit = domain == TrackerDomain.Udp
? _options.UdpTrackedKeyLimit
: _options.MaxTrackedKeys - _options.UdpTrackedKeyLimit;
int maxTrackedKeys = domain == TrackerDomain.Udp || canUseCriticalReserve
? partitionLimit
: partitionLimit - _options.CriticalTrackedKeyReserve;
if (!CanAcquireAll(
tracker,
tracker.WindowCounts,
rates,
maxTrackedKeys,
ref stagedNewKeys)
|| !CanAcquireAll(
tracker,
tracker.ConcurrencyCounts,
concurrency,
maxTrackedKeys,
ref stagedNewKeys))
{
lease = null;
return false;
}
foreach (RateDimension dimension in rates)
{
tracker.WindowCounts[dimension.Key] =
tracker.WindowCounts.GetValueOrDefault(dimension.Key) + 1;
}
if (concurrency.IsEmpty)
{
lease = null;
return true;
}
string[] acquiredConcurrency = new string[concurrency.Length];
for (int index = 0; index < concurrency.Length; index++)
{
RateDimension dimension = concurrency[index];
tracker.ConcurrencyCounts[dimension.Key] =
tracker.ConcurrencyCounts.GetValueOrDefault(dimension.Key) + 1;
acquiredConcurrency[index] = dimension.Key;
}
lease = new AbuseLease(this, tracker, acquiredConcurrency);
return true;
}
private static bool CanAcquireAll(
TrackerState tracker,
Dictionary<string, int> counts,
ReadOnlySpan<RateDimension> dimensions,
int maxTrackedKeys,
ref int stagedNewKeys)
{
foreach (RateDimension dimension in dimensions)
{
if (counts.TryGetValue(dimension.Key, out int current))
{
if (current >= dimension.Limit)
{
return false;
}
continue;
}
stagedNewKeys++;
if (tracker.WindowCounts.Count + tracker.ConcurrencyCounts.Count + stagedNewKeys
> maxTrackedKeys)
{
return false;
}
}
return true;
}
private static void Release(TrackerState tracker, string[] keys)
{
lock (tracker.Gate)
{
foreach (string key in keys)
{
if (!tracker.ConcurrencyCounts.TryGetValue(key, out int current))
{
continue;
}
if (current <= 1)
{
tracker.ConcurrencyCounts.Remove(key);
}
else
{
tracker.ConcurrencyCounts[key] = current - 1;
}
}
}
}
private static bool DisposeAccepted(AbuseLease? lease)
{
lease?.Dispose();
return true;
}
private static bool IsLeaseCriticalOperation(string operation) => operation is
"RenewSessionLease" or "UpdateSession" or "DeleteSession";
private static string GetNetworkPrefix(IPAddress? address)
{
if (address is null)
{
return "unknown";
}
IPAddress normalized = address.IsIPv4MappedToIPv6 ? address.MapToIPv4() : address;
byte[] bytes = normalized.GetAddressBytes();
if (bytes.Length == 4)
{
bytes[3] = 0;
return $"4:{Convert.ToHexString(bytes)}:24";
}
if (bytes.Length == 16)
{
Array.Clear(bytes, 7, 9);
return $"6:{Convert.ToHexString(bytes)}:56";
}
return "unknown";
}
private static IPAddress NormalizeAddress(IPAddress address) =>
address.IsIPv4MappedToIPv6 ? address.MapToIPv4() : address;
private readonly record struct RateDimension(string Key, int Limit);
private enum TrackerDomain
{
Http,
Udp,
}
internal sealed class TrackerState(DateTimeOffset windowStartedAt)
{
public object Gate { get; } = new();
public Dictionary<string, int> WindowCounts { get; } = new(StringComparer.Ordinal);
public Dictionary<string, int> ConcurrencyCounts { get; } = new(StringComparer.Ordinal);
public DateTimeOffset WindowStartedAt { get; set; } = windowStartedAt;
}
internal sealed class AbuseLease : IDisposable
{
private AbuseProtectionService? _owner;
private readonly TrackerState _tracker;
private readonly string[] _keys;
internal AbuseLease(
AbuseProtectionService owner,
TrackerState tracker,
string[] keys)
{
_owner = owner;
_tracker = tracker;
_keys = keys;
}
public void Dispose()
{
if (Interlocked.Exchange(ref _owner, null) is not null)
{
Release(_tracker, _keys);
}
}
}
}
@@ -0,0 +1,135 @@
using FinalFactory.Rendezvous.Contracts;
using Microsoft.AspNetCore.Http.Features;
namespace FinalFactory.Rendezvous.Server.Abuse;
internal sealed class HttpAbuseProtectionMiddleware(
RequestDelegate next,
AbuseProtectionService protection)
{
public async Task InvokeAsync(HttpContext context)
{
IHttpMaxRequestBodySizeFeature? bodySize =
context.Features.Get<IHttpMaxRequestBodySizeFeature>();
if (bodySize is { IsReadOnly: false })
{
bodySize.MaxRequestBodySize = ContractLimits.HttpRequestMaxBytes;
}
string operation = context.GetEndpoint()?.Metadata.GetMetadata<IEndpointNameMetadata>()
?.EndpointName ?? "Unmatched";
bool healthEndpoint = operation is "GetLiveness" or "GetReadiness";
bool operatorEndpoint = operation is
"GetOperatorStatus"
or "RevokeOperatorListing"
or "RevokeOperatorPrincipal"
or "RevokeOperatorSigningKey"
or "BeginOperatorDrain";
if (operatorEndpoint
&& !protection.IsOperatorSourceAllowed(context.Connection.RemoteIpAddress))
{
bool deniedSourceAdmitted = protection.TryAcquireHttpIngress(
context.Connection.RemoteIpAddress,
"Unmatched",
out AbuseProtectionService.AbuseLease? deniedSourceLease,
out int deniedRetryAfterSeconds);
using (deniedSourceLease)
{
if (!deniedSourceAdmitted)
{
context.Response.Headers.RetryAfter = deniedRetryAfterSeconds.ToString(
System.Globalization.CultureInfo.InvariantCulture);
await WriteErrorAsync(
context,
StatusCodes.Status429TooManyRequests,
RendezvousErrorCode.RateLimited,
"The request rate limit was exceeded.",
deniedRetryAfterSeconds).ConfigureAwait(false);
return;
}
await WriteErrorAsync(
context,
StatusCodes.Status404NotFound,
RendezvousErrorCode.NotFound,
"The requested resource was not found.").ConfigureAwait(false);
}
return;
}
AbuseProtectionService.AbuseLease? lease;
int retryAfterSeconds;
bool acquired;
if (healthEndpoint)
{
acquired = protection.TryAcquireHealthIngress(
context.Connection.RemoteIpAddress,
out lease,
out retryAfterSeconds);
}
else if (operatorEndpoint)
{
acquired = protection.TryAcquireOperatorIngress(
context.Connection.RemoteIpAddress,
out lease,
out retryAfterSeconds);
}
else
{
acquired = protection.TryAcquireHttpIngress(
context.Connection.RemoteIpAddress,
operation,
out lease,
out retryAfterSeconds);
}
if (!acquired)
{
context.Response.Headers.RetryAfter = retryAfterSeconds.ToString(
System.Globalization.CultureInfo.InvariantCulture);
await WriteErrorAsync(
context,
StatusCodes.Status429TooManyRequests,
RendezvousErrorCode.RateLimited,
"The request rate limit was exceeded.",
retryAfterSeconds).ConfigureAwait(false);
return;
}
using (lease)
{
if (context.Request.ContentLength > ContractLimits.HttpRequestMaxBytes)
{
await WriteErrorAsync(
context,
StatusCodes.Status413PayloadTooLarge,
RendezvousErrorCode.InvalidRequest,
"The request body exceeds the supported size.").ConfigureAwait(false);
return;
}
await next(context).ConfigureAwait(false);
}
}
private static Task WriteErrorAsync(
HttpContext context,
int status,
RendezvousErrorCode code,
string message,
int? retryAfterSeconds = null)
{
context.Response.StatusCode = status;
return context.Response.WriteAsJsonAsync(
new ApiError
{
Code = code,
Message = message,
RetryAfterSeconds = retryAfterSeconds,
},
ContractJson.Options,
contentType: "application/json",
cancellationToken: context.RequestAborted);
}
}
@@ -0,0 +1,24 @@
using System.Net;
using Microsoft.AspNetCore.HttpOverrides;
namespace FinalFactory.Rendezvous.Server.Abuse;
internal static class TrustedProxyForwarding
{
public static bool IsEnabled(AbuseProtectionOptions options) =>
options.TrustedProxyAddresses is { Length: > 0 };
public static void Configure(
ForwardedHeadersOptions forwarded,
AbuseProtectionOptions abuse)
{
forwarded.ForwardedHeaders = ForwardedHeaders.XForwardedFor;
forwarded.ForwardLimit = 1;
forwarded.KnownProxies.Clear();
forwarded.KnownIPNetworks.Clear();
foreach (string address in abuse.TrustedProxyAddresses ?? [])
{
forwarded.KnownProxies.Add(IPAddress.Parse(address));
}
}
}
@@ -1,4 +1,5 @@
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.Sessions;
using FinalFactory.Rendezvous.Server.State;
@@ -15,6 +16,10 @@ internal sealed class ConnectionOutcomeMetrics
{
private readonly object _gate = new();
private readonly Dictionary<(ConnectionOutcomeKind, ConnectionElapsedBucket), long> _counts = [];
private readonly RendezvousTelemetry? _telemetry;
public ConnectionOutcomeMetrics(RendezvousTelemetry? telemetry = null) =>
_telemetry = telemetry;
internal void Record(ConnectionOutcomeKind outcome, ConnectionElapsedBucket elapsedBucket)
{
@@ -24,6 +29,8 @@ internal sealed class ConnectionOutcomeMetrics
_counts.TryGetValue(key, out long count);
_counts[key] = count + 1;
}
_telemetry?.RecordConnectionOutcome(outcome.ToString(), elapsedBucket.ToString());
}
internal long GetCount(ConnectionOutcomeKind outcome, ConnectionElapsedBucket elapsedBucket)
@@ -0,0 +1,235 @@
using System.ComponentModel.DataAnnotations;
using System.Net;
using FinalFactory.Rendezvous.Server.Abuse;
namespace FinalFactory.Rendezvous.Server.Deployment;
internal sealed record DeploymentOptions
{
public const string SectionName = "Rendezvous:Deployment";
[Required]
public string PublicHttpBaseUrl { get; init; } = string.Empty;
[Required]
public string PublicUdpHost { get; init; } = string.Empty;
[Range(1, 65_535)]
public int PublicUdpPort { get; init; } = 9050;
[Range(1, 30)]
public int DrainDeadlineSeconds { get; init; } = 30;
[Range(0, 5)]
public int MinimumDrainSeconds { get; init; } = 1;
public bool SingleActiveInstance { get; init; } = true;
public bool AllowPrivatePublicEndpoints { get; init; }
public IReadOnlyList<string> ValidateProduction(
AbuseProtectionOptions abuseProtection,
string? allowedHosts)
{
List<string> errors = [];
if (!SingleActiveInstance)
{
errors.Add("Rendezvous:Deployment:SingleActiveInstance must be true because ephemeral state is not shared between replicas.");
}
if (MinimumDrainSeconds >= DrainDeadlineSeconds)
{
errors.Add("Rendezvous:Deployment:MinimumDrainSeconds must be less than DrainDeadlineSeconds.");
}
if (DrainDeadlineSeconds is < 1 or > 30)
{
errors.Add("Rendezvous:Deployment:DrainDeadlineSeconds must be between 1 and 30.");
}
if (MinimumDrainSeconds is < 0 or > 5)
{
errors.Add("Rendezvous:Deployment:MinimumDrainSeconds must be between 0 and 5.");
}
if (PublicUdpPort is < 1 or > 65_535)
{
errors.Add("Rendezvous:Deployment:PublicUdpPort must be between 1 and 65535.");
}
ValidateHttpEndpoint(errors);
ValidateUdpEndpoint(errors);
ValidateAllowedHosts(errors, allowedHosts, PublicHttpBaseUrl);
if (abuseProtection.TrustedProxyAddresses is not { Length: > 0 })
{
errors.Add(
"Rendezvous:AbuseProtection:TrustedProxyAddresses must list the exact TLS proxy addresses; forwarded headers are rejected without this trust boundary.");
}
return errors;
}
private void ValidateHttpEndpoint(List<string> errors)
{
if (!Uri.TryCreate(PublicHttpBaseUrl, UriKind.Absolute, out Uri? endpoint)
|| !string.Equals(endpoint.Scheme, Uri.UriSchemeHttps, StringComparison.Ordinal)
|| !string.IsNullOrEmpty(endpoint.UserInfo)
|| !string.IsNullOrEmpty(endpoint.Query)
|| !string.IsNullOrEmpty(endpoint.Fragment)
|| endpoint.AbsolutePath != "/")
{
errors.Add(
"Rendezvous:Deployment:PublicHttpBaseUrl must be an absolute HTTPS origin with no credentials, path, query, or fragment (for example, https://rendezvous.your-company.tld/).");
return;
}
if (!AllowPrivatePublicEndpoints && !IsPublicHost(endpoint.Host))
{
errors.Add(
"Rendezvous:Deployment:PublicHttpBaseUrl must use a public DNS name or address; set AllowPrivatePublicEndpoints=true only for an isolated deployment smoke test.");
}
}
private void ValidateUdpEndpoint(List<string> errors)
{
if (string.IsNullOrWhiteSpace(PublicUdpHost)
|| PublicUdpHost.Contains("//", StringComparison.Ordinal)
|| PublicUdpHost.Contains(':', StringComparison.Ordinal) && !IPAddress.TryParse(PublicUdpHost, out _)
|| Uri.CheckHostName(PublicUdpHost) == UriHostNameType.Unknown)
{
errors.Add(
"Rendezvous:Deployment:PublicUdpHost must contain only the advertised DNS name or IP address; configure the port separately.");
return;
}
if (!AllowPrivatePublicEndpoints && !IsPublicHost(PublicUdpHost))
{
errors.Add(
"Rendezvous:Deployment:PublicUdpHost must use a public DNS name or address; set AllowPrivatePublicEndpoints=true only for an isolated deployment smoke test.");
}
}
private static void ValidateAllowedHosts(
List<string> errors,
string? allowedHosts,
string publicHttpBaseUrl)
{
string[] hosts = (allowedHosts ?? string.Empty).Split(
';',
StringSplitOptions.RemoveEmptyEntries | StringSplitOptions.TrimEntries);
if (hosts.Length == 0 || hosts.Any(static host => host is "*" or "+"))
{
errors.Add(
"AllowedHosts must explicitly list the public HTTP host in production; wildcard or empty host filtering is unsafe.");
return;
}
if (Uri.TryCreate(publicHttpBaseUrl, UriKind.Absolute, out Uri? endpoint)
&& !hosts.Contains(endpoint.Host, StringComparer.OrdinalIgnoreCase))
{
errors.Add(
"AllowedHosts must contain the exact host advertised by Rendezvous:Deployment:PublicHttpBaseUrl.");
}
}
private static bool IsPublicHost(string host)
{
if (!IPAddress.TryParse(host, out IPAddress? address))
{
return Uri.CheckHostName(host) == UriHostNameType.Dns
&& host.Contains('.', StringComparison.Ordinal)
&& !IsReservedDnsName(host);
}
return IsGloballyRoutableUnicast(address);
}
private static bool IsReservedDnsName(string host)
{
string normalized = host.TrimEnd('.');
string[] reservedSuffixes =
[
"localhost",
"local",
"invalid",
"test",
"example",
"example.com",
"example.net",
"example.org",
"home.arpa",
"alt",
"onion",
];
return reservedSuffixes.Any(suffix =>
string.Equals(normalized, suffix, StringComparison.OrdinalIgnoreCase)
|| normalized.EndsWith($".{suffix}", StringComparison.OrdinalIgnoreCase));
}
private static bool IsGloballyRoutableUnicast(IPAddress address)
{
if (address.IsIPv4MappedToIPv6)
{
address = address.MapToIPv4();
}
byte[] bytes = address.GetAddressBytes();
if (address.AddressFamily == System.Net.Sockets.AddressFamily.InterNetwork)
{
return !(bytes[0] is 0 or 10 or 127
|| bytes[0] == 100 && bytes[1] is >= 64 and <= 127
|| bytes[0] == 169 && bytes[1] == 254
|| bytes[0] == 172 && bytes[1] is >= 16 and <= 31
|| bytes[0] == 192
&& (bytes[1] == 0 && bytes[2] is 0 or 2
|| bytes[1] == 88 && bytes[2] == 99
|| bytes[1] == 168)
|| bytes[0] == 198
&& (bytes[1] is 18 or 19
|| bytes[1] == 51 && bytes[2] == 100)
|| bytes[0] == 203 && bytes[1] == 0 && bytes[2] == 113
|| bytes[0] >= 224);
}
return address.AddressFamily == System.Net.Sockets.AddressFamily.InterNetworkV6
&& !IPAddress.IsLoopback(address)
&& !address.Equals(IPAddress.IPv6Any)
&& !address.IsIPv6LinkLocal
&& !address.IsIPv6SiteLocal
&& !address.IsIPv6Multicast
&& (bytes[0] & 0xe0) == 0x20
&& !HasPrefix(bytes, [0x20, 0x01, 0x00], 23)
&& !HasPrefix(bytes, [0x20, 0x01, 0x0d, 0xb8], 32)
&& !HasPrefix(bytes, [0x20, 0x02], 16)
&& !HasPrefix(bytes, [0x3f, 0xff, 0x00], 20);
}
private static bool HasPrefix(byte[] address, byte[] prefix, int bitCount)
{
int fullBytes = bitCount / 8;
for (int index = 0; index < fullBytes; index++)
{
if (address[index] != prefix[index])
{
return false;
}
}
int remainingBits = bitCount % 8;
if (remainingBits == 0)
{
return true;
}
int mask = 0xff << (8 - remainingBits);
return (address[fullBytes] & mask) == (prefix[fullBytes] & mask);
}
}
internal sealed class DeploymentConfigurationException(IReadOnlyList<string> errors)
: InvalidOperationException(
"Production deployment configuration is invalid:" + Environment.NewLine
+ string.Join(Environment.NewLine, errors.Select(static error => $"- {error}")))
{
}
@@ -0,0 +1,102 @@
using System.Diagnostics;
using FinalFactory.Rendezvous.Server.State;
using Microsoft.Extensions.Options;
namespace FinalFactory.Rendezvous.Server.Deployment;
internal sealed class GracefulDrainService : IHostedService, IDisposable
{
private static readonly TimeSpan PollInterval = TimeSpan.FromMilliseconds(50);
private readonly InMemoryEphemeralRendezvousStore _store;
private readonly IHostApplicationLifetime _lifetime;
private readonly DeploymentOptions _options;
private readonly ILogger<GracefulDrainService> _logger;
private readonly object _gate = new();
private CancellationTokenRegistration _stoppingRegistration;
private Task? _drainTask;
public GracefulDrainService(
InMemoryEphemeralRendezvousStore store,
IHostApplicationLifetime lifetime,
IOptions<DeploymentOptions> options,
ILogger<GracefulDrainService> logger)
{
_store = store;
_lifetime = lifetime;
_options = options.Value;
_logger = logger;
}
public Task StartAsync(CancellationToken cancellationToken)
{
cancellationToken.ThrowIfCancellationRequested();
_stoppingRegistration = _lifetime.ApplicationStopping.Register(
() => EnsureDrainAsync().GetAwaiter().GetResult());
return Task.CompletedTask;
}
public Task StopAsync(CancellationToken cancellationToken)
{
// ApplicationStopping callbacks run before hosted services and listeners
// stop. StopAsync is the idempotent fallback for directly driven hosts.
_ = cancellationToken;
return EnsureDrainAsync();
}
public void Dispose() => _stoppingRegistration.Dispose();
private Task EnsureDrainAsync()
{
lock (_gate)
{
return _drainTask ??= DrainAsync();
}
}
private async Task DrainAsync()
{
_store.BeginDrain(CancellationToken.None);
TimeSpan deadline = TimeSpan.FromSeconds(_options.DrainDeadlineSeconds);
TimeSpan minimum = TimeSpan.FromSeconds(_options.MinimumDrainSeconds);
long startedAt = Stopwatch.GetTimestamp();
LogDrainStarted(_logger, _options.DrainDeadlineSeconds);
try
{
while (Stopwatch.GetElapsedTime(startedAt) < deadline)
{
TimeSpan elapsed = Stopwatch.GetElapsedTime(startedAt);
if (elapsed >= minimum && _store.GetActiveJoinAttemptCountForDrain() == 0)
{
break;
}
TimeSpan remaining = deadline - elapsed;
await Task.Delay(
remaining < PollInterval ? remaining : PollInterval,
CancellationToken.None).ConfigureAwait(false);
}
}
finally
{
_store.MarkUnavailable();
double elapsedMilliseconds = Stopwatch.GetElapsedTime(startedAt).TotalMilliseconds;
LogDrainFinished(_logger, elapsedMilliseconds);
}
}
private static readonly Action<ILogger, int, Exception?> DrainStarted = LoggerMessage.Define<int>(
LogLevel.Information,
new EventId(1, nameof(LogDrainStarted)),
"Graceful drain started with a {DrainDeadlineSeconds}-second deadline");
private static readonly Action<ILogger, double, Exception?> DrainFinished = LoggerMessage.Define<double>(
LogLevel.Information,
new EventId(2, nameof(LogDrainFinished)),
"Graceful drain finished after {ElapsedMilliseconds:F0} ms; ephemeral state was cleared");
private static void LogDrainStarted(ILogger logger, int drainDeadlineSeconds) =>
DrainStarted(logger, drainDeadlineSeconds, null);
private static void LogDrainFinished(ILogger logger, double elapsedMilliseconds) =>
DrainFinished(logger, elapsedMilliseconds, null);
}
@@ -3,6 +3,7 @@
<TargetFramework>net10.0</TargetFramework>
<AssemblyName>FinalFactory.Rendezvous.Server</AssemblyName>
<RootNamespace>FinalFactory.Rendezvous.Server</RootNamespace>
<EnableDefaultCompileItems>false</EnableDefaultCompileItems>
<IsPackable>false</IsPackable>
<OpenApiGenerateDocuments>true</OpenApiGenerateDocuments>
<OpenApiDocumentsDirectory>$(MSBuildProjectDirectory)/../../docs/api</OpenApiDocumentsDirectory>
@@ -14,4 +15,74 @@
<PackageReference Include="Microsoft.AspNetCore.OpenApi" />
<PackageReference Include="Microsoft.Extensions.ApiDescription.Server" PrivateAssets="all" />
</ItemGroup>
<ItemGroup>
<AssemblyAttribute Include="System.Reflection.AssemblyMetadataAttribute">
<_Parameter1>RendezvousMinimumClientVersion</_Parameter1>
<_Parameter2>$(MinimumClientVersion)</_Parameter2>
</AssemblyAttribute>
<AssemblyAttribute Include="System.Reflection.AssemblyMetadataAttribute">
<_Parameter1>RendezvousMaximumClientMajorVersion</_Parameter1>
<_Parameter2>$(MaximumClientMajorVersion)</_Parameter2>
</AssemblyAttribute>
<AssemblyAttribute Include="System.Reflection.AssemblyMetadataAttribute">
<_Parameter1>RendezvousUdpContractVersion</_Parameter1>
<_Parameter2>$(UdpContractVersion)</_Parameter2>
</AssemblyAttribute>
<AssemblyAttribute Include="System.Reflection.AssemblyMetadataAttribute">
<_Parameter1>RendezvousConnectionTicketFormatVersion</_Parameter1>
<_Parameter2>$(ConnectionTicketFormatVersion)</_Parameter2>
</AssemblyAttribute>
<AssemblyAttribute Include="System.Reflection.AssemblyMetadataAttribute">
<_Parameter1>RendezvousLiteNetLibMajorVersion</_Parameter1>
<_Parameter2>$(LiteNetLibMajorVersion)</_Parameter2>
</AssemblyAttribute>
</ItemGroup>
<ItemGroup>
<!-- Roslyn and source generators consume syntax trees in item order. Keep
this ordinal manifest explicit so clean builds are byte reproducible. -->
<Compile Include="Abuse/AbuseProtectionOptions.cs" />
<Compile Include="Abuse/AbuseProtectionService.cs" />
<Compile Include="Abuse/HttpAbuseProtectionMiddleware.cs" />
<Compile Include="Abuse/TrustedProxyForwarding.cs" />
<Compile Include="Browser/EphemeralCursorProtector.cs" />
<Compile Include="Browser/SessionBrowserCursorCodec.cs" />
<Compile Include="Browser/SessionBrowserService.cs" />
<Compile Include="ConnectionOutcomes/ConnectionOutcomeService.cs" />
<Compile Include="Deployment/DeploymentOptions.cs" />
<Compile Include="Deployment/GracefulDrainService.cs" />
<Compile Include="Http/ContractEndpoints.cs" />
<Compile Include="Http/RendezvousExceptionHandler.cs" />
<Compile Include="JoinAttempts/JoinAttemptCursorCodec.cs" />
<Compile Include="JoinAttempts/JoinAttemptService.cs" />
<Compile Include="Observability/AuditOptions.cs" />
<Compile Include="Observability/AuditTrail.cs" />
<Compile Include="Observability/HealthEndpoints.cs" />
<Compile Include="Observability/RendezvousReadiness.cs" />
<Compile Include="Observability/RendezvousTelemetry.cs" />
<Compile Include="Observability/TelemetryMiddleware.cs" />
<Compile Include="Operations/OperatorEndpoints.cs" />
<Compile Include="Operations/OperatorModels.cs" />
<Compile Include="Operations/OperatorService.cs" />
<Compile Include="Operations/ReleaseCompatibility.cs" />
<Compile Include="Program.cs" />
<Compile Include="Properties/AssemblyInfo.cs" />
<Compile Include="Provisioning/GamePolicy.cs" />
<Compile Include="Provisioning/GamePolicyRegistry.cs" />
<Compile Include="Provisioning/PrincipalCredentialService.cs" />
<Compile Include="Provisioning/Principals.cs" />
<Compile Include="Provisioning/ProvisioningOptions.cs" />
<Compile Include="Provisioning/ProvisioningRuntime.cs" />
<Compile Include="Provisioning/PublisherAuthorizationService.cs" />
<Compile Include="Provisioning/SecretProviders.cs" />
<Compile Include="Provisioning/SigningKeyRing.cs" />
<Compile Include="Sessions/EphemeralCapabilityIssuer.cs" />
<Compile Include="Sessions/SessionLeaseService.cs" />
<Compile Include="State/EphemeralStateContracts.cs" />
<Compile Include="State/InMemoryEphemeralRendezvousStore.cs" />
<Compile Include="State/StoreResultMapping.cs" />
<Compile Include="Transport/LiteNetNatRequestCodec.cs" />
<Compile Include="Transport/NatMediationProcessor.cs" />
<Compile Include="Transport/UdpMediatorOptions.cs" />
<Compile Include="Transport/UdpMediatorService.cs" />
</ItemGroup>
</Project>
@@ -1,5 +1,6 @@
using System.Net;
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Abuse;
using FinalFactory.Rendezvous.Server.Browser;
using FinalFactory.Rendezvous.Server.ConnectionOutcomes;
using FinalFactory.Rendezvous.Server.JoinAttempts;
@@ -20,6 +21,7 @@ internal static class ContractEndpoints
.Accepts<RegisterSessionRequest>("application/json")
.Produces<RegisterSessionResponse>(StatusCodes.Status201Created)
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status409Conflict)
@@ -31,45 +33,54 @@ internal static class ContractEndpoints
.Accepts<RenewLeaseRequest>("application/json")
.Produces<RenewLeaseResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status409Conflict)
.Produces<ApiError>(StatusCodes.Status410Gone)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("RenewSessionLease");
sessions.MapPut("/{listingId}", UpdateSession)
.Accepts<UpdateSessionRequest>("application/json")
.Produces(StatusCodes.Status204NoContent)
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("UpdateSession");
sessions.MapDelete("/{listingId}", DeleteSession)
.Accepts<DeleteSessionRequest>("application/json")
.Produces(StatusCodes.Status204NoContent)
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("DeleteSession");
sessions.MapGet("/", BrowseSessions)
.Produces<BrowseSessionsResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("BrowseSessions");
sessions.MapGet("/{listingId}", GetSession)
.Produces<GetSessionResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("GetSession");
sessions.MapGet("/{listingId}/join-attempts", BrowseHostJoinAttempts)
.Produces<BrowseHostJoinAttemptsResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("BrowseHostJoinAttempts");
@@ -80,6 +91,7 @@ internal static class ContractEndpoints
.Accepts<CreateJoinAttemptRequest>("application/json")
.Produces<CreateJoinAttemptResponse>(StatusCodes.Status201Created)
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status409Conflict)
.Produces<ApiError>(StatusCodes.Status410Gone)
@@ -90,14 +102,17 @@ internal static class ContractEndpoints
.Produces(StatusCodes.Status204NoContent)
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("CancelJoinAttempt");
attempts.MapPost("/{attemptId}/outcome", ReportConnectionOutcome)
.Accepts<ReportConnectionOutcomeRequest>("application/json")
.Produces<ReportConnectionOutcomeResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status409Conflict)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("ReportConnectionOutcome");
@@ -109,6 +124,7 @@ internal static class ContractEndpoints
[FromHeader(Name = "Authorization")] string? authorizationHeader,
[FromServices] PrincipalCredentialService credentials,
[FromServices] SessionLeaseService sessions,
[FromServices] AbuseProtectionService abuseProtection,
[FromServices] IWallClock clock,
HttpContext httpContext,
CancellationToken cancellationToken)
@@ -122,13 +138,29 @@ internal static class ContractEndpoints
return AuthenticationRequired(httpContext);
}
SessionServiceResult<RegisterSessionResponse> result = sessions.Register(
principal!,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Created($"/v1/sessions/{result.Value.ListingId}", result.Value)
: Error(result.Error);
IPublisherPrincipal publisher = (IPublisherPrincipal)principal!;
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"RegisterSession",
Tenant(publisher.GameId, publisher.EnvironmentId),
publisher.Subject,
null,
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
SessionServiceResult<RegisterSessionResponse> result = sessions.Register(
principal!,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Created($"/v1/sessions/{result.Value.ListingId}", result.Value)
: Error(result.Error);
}
}
private static IResult RenewLease(
@@ -137,6 +169,7 @@ internal static class ContractEndpoints
[FromHeader(Name = "Authorization")] string? authorizationHeader,
[FromServices] PrincipalCredentialService credentials,
[FromServices] SessionLeaseService sessions,
[FromServices] AbuseProtectionService abuseProtection,
[FromServices] IWallClock clock,
HttpContext httpContext,
CancellationToken cancellationToken)
@@ -150,14 +183,30 @@ internal static class ContractEndpoints
return AuthenticationRequired(httpContext);
}
SessionServiceResult<RenewLeaseResponse> result = sessions.Renew(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
IPublisherPrincipal publisher = (IPublisherPrincipal)principal!;
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"RenewSessionLease",
Tenant(publisher.GameId, publisher.EnvironmentId),
publisher.Subject,
listingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
SessionServiceResult<RenewLeaseResponse> result = sessions.Renew(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
}
}
private static IResult UpdateSession(
@@ -166,6 +215,7 @@ internal static class ContractEndpoints
[FromHeader(Name = "Authorization")] string? authorizationHeader,
[FromServices] PrincipalCredentialService credentials,
[FromServices] SessionLeaseService sessions,
[FromServices] AbuseProtectionService abuseProtection,
[FromServices] IWallClock clock,
HttpContext httpContext,
CancellationToken cancellationToken)
@@ -179,12 +229,28 @@ internal static class ContractEndpoints
return AuthenticationRequired(httpContext);
}
SessionServiceResult<bool> result = sessions.Update(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
IPublisherPrincipal publisher = (IPublisherPrincipal)principal!;
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"UpdateSession",
Tenant(publisher.GameId, publisher.EnvironmentId),
publisher.Subject,
listingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
SessionServiceResult<bool> result = sessions.Update(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
}
}
private static IResult DeleteSession(
@@ -193,6 +259,7 @@ internal static class ContractEndpoints
[FromHeader(Name = "Authorization")] string? authorizationHeader,
[FromServices] PrincipalCredentialService credentials,
[FromServices] SessionLeaseService sessions,
[FromServices] AbuseProtectionService abuseProtection,
[FromServices] IWallClock clock,
HttpContext httpContext,
CancellationToken cancellationToken)
@@ -206,12 +273,28 @@ internal static class ContractEndpoints
return AuthenticationRequired(httpContext);
}
SessionServiceResult<bool> result = sessions.Delete(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
IPublisherPrincipal publisher = (IPublisherPrincipal)principal!;
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"DeleteSession",
Tenant(publisher.GameId, publisher.EnvironmentId),
publisher.Subject,
listingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
SessionServiceResult<bool> result = sessions.Delete(
principal!,
listingId,
request,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
}
}
private static IResult BrowseSessions(
@@ -224,6 +307,8 @@ internal static class ContractEndpoints
[FromQuery] bool? excludeFull,
[FromQuery] string? cursor,
[FromServices] SessionBrowserService browser,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
if (!GameId.TryParse(gameId, out GameId parsedGameId)
@@ -233,20 +318,35 @@ internal static class ContractEndpoints
return Error(RendezvousErrorCode.InvalidRequest);
}
BrowserServiceResult<BrowseSessionsResponse> result = browser.Browse(new()
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"BrowseSessions",
Tenant(parsedGameId, parsedEnvironmentId),
null,
null,
out AbuseProtectionService.AbuseLease? abuseLease))
{
ContractVersion = contractVersion,
GameId = parsedGameId,
EnvironmentId = parsedEnvironmentId,
ProtocolVersion = protocolVersion,
RegionId = regionId is null ? null : new RegionId(regionId),
PageSize = pageSize ?? ContractLimits.BrowserPageMaxItems,
ExcludeFull = excludeFull ?? false,
Cursor = cursor,
}, cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
return RateLimited(httpContext);
}
using (abuseLease)
{
BrowserServiceResult<BrowseSessionsResponse> result = browser.Browse(new()
{
ContractVersion = contractVersion,
GameId = parsedGameId,
EnvironmentId = parsedEnvironmentId,
ProtocolVersion = protocolVersion,
RegionId = regionId is null ? null : new RegionId(regionId),
PageSize = pageSize ?? ContractLimits.BrowserPageMaxItems,
ExcludeFull = excludeFull ?? false,
Cursor = cursor,
}, cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
}
}
private static IResult GetSession(
@@ -256,6 +356,8 @@ internal static class ContractEndpoints
[FromQuery] string environmentId,
[FromQuery] uint protocolVersion,
[FromServices] SessionBrowserService browser,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
if (ContractValidation.ValidateContractVersion(contractVersion) != RendezvousErrorCode.None)
@@ -269,15 +371,30 @@ internal static class ContractEndpoints
return Error(RendezvousErrorCode.InvalidRequest);
}
BrowserServiceResult<GetSessionResponse> result = browser.Get(
listingId,
parsedGameId,
parsedEnvironmentId,
protocolVersion,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"GetSession",
Tenant(parsedGameId, parsedEnvironmentId),
null,
listingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
BrowserServiceResult<GetSessionResponse> result = browser.Get(
listingId,
parsedGameId,
parsedEnvironmentId,
protocolVersion,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
}
}
private static IResult BrowseHostJoinAttempts(
@@ -287,23 +404,41 @@ internal static class ContractEndpoints
[FromQuery] int? pageSize,
[FromQuery] string? cursor,
[FromServices] JoinAttemptService attempts,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
JoinAttemptServiceResult<BrowseHostJoinAttemptsResponse> result = attempts.BrowseForHost(
listingId,
contractVersion,
leaseToken,
pageSize ?? ContractLimits.BrowserPageMaxItems,
cursor,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"BrowseHostJoinAttempts",
null,
AbuseProtectionService.FingerprintSecret(leaseToken ?? string.Empty),
listingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
JoinAttemptServiceResult<BrowseHostJoinAttemptsResponse> result = attempts.BrowseForHost(
listingId,
contractVersion,
leaseToken,
pageSize ?? ContractLimits.BrowserPageMaxItems,
cursor,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
}
}
private static IResult CreateJoinAttempt(
[FromBody] CreateJoinAttemptRequest request,
[FromServices] JoinAttemptService attempts,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
@@ -313,26 +448,58 @@ internal static class ContractEndpoints
}
string clientSubject = attempts.CreateAnonymousClientSubject(remoteAddress);
JoinAttemptServiceResult<CreateJoinAttemptResponse> result = attempts.Create(
clientSubject,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Created($"/v1/join-attempts/{result.Value.AttemptId}", result.Value)
: Error(result.Error);
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"CreateJoinAttempt",
Tenant(request.GameId, request.EnvironmentId),
clientSubject,
request.ListingId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
JoinAttemptServiceResult<CreateJoinAttemptResponse> result = attempts.Create(
clientSubject,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Created($"/v1/join-attempts/{result.Value.AttemptId}", result.Value)
: Error(result.Error);
}
}
private static IResult CancelJoinAttempt(
JoinAttemptId attemptId,
[FromHeader(Name = "X-Rendezvous-Client-Punch-Capability")] string clientPunchCapability,
[FromServices] JoinAttemptService attempts,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
JoinAttemptServiceResult<bool> result = attempts.Cancel(
attemptId,
clientPunchCapability,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"CancelJoinAttempt",
null,
AbuseProtectionService.FingerprintSecret(clientPunchCapability ?? string.Empty),
attemptId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
JoinAttemptServiceResult<bool> result = attempts.Cancel(
attemptId,
clientPunchCapability,
cancellationToken);
return result.Succeeded ? Results.NoContent() : Error(result.Error);
}
}
private static IResult ReportConnectionOutcome(
@@ -340,16 +507,33 @@ internal static class ContractEndpoints
[FromHeader(Name = "X-Rendezvous-Client-Punch-Capability")] string clientPunchCapability,
[FromBody] ReportConnectionOutcomeRequest request,
[FromServices] ConnectionOutcomeService outcomes,
[FromServices] AbuseProtectionService abuseProtection,
HttpContext httpContext,
CancellationToken cancellationToken)
{
ConnectionOutcomeServiceResult result = outcomes.Report(
attemptId,
clientPunchCapability,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
if (!TryAcquireIdentity(
abuseProtection,
httpContext,
"ReportConnectionOutcome",
null,
AbuseProtectionService.FingerprintSecret(clientPunchCapability ?? string.Empty),
attemptId.ToString(),
out AbuseProtectionService.AbuseLease? abuseLease))
{
return RateLimited(httpContext);
}
using (abuseLease)
{
ConnectionOutcomeServiceResult result = outcomes.Report(
attemptId,
clientPunchCapability,
request,
cancellationToken);
return result.Succeeded && result.Value is not null
? Results.Ok(result.Value)
: Error(result.Error);
}
}
private static bool TryAuthenticatePublisher(
@@ -377,11 +561,41 @@ internal static class ContractEndpoints
return true;
}
private static IResult Error(RendezvousErrorCode code) => Results.Json(
private static bool TryAcquireIdentity(
AbuseProtectionService abuseProtection,
HttpContext httpContext,
string operation,
string? tenant,
string? principal,
string? resource,
out AbuseProtectionService.AbuseLease? lease)
{
if (abuseProtection.TryAcquireHttpIdentity(
operation,
httpContext.Connection.RemoteIpAddress,
tenant,
principal,
resource,
out lease,
out int retryAfterSeconds))
{
return true;
}
httpContext.Response.Headers.RetryAfter = retryAfterSeconds.ToString(
System.Globalization.CultureInfo.InvariantCulture);
return false;
}
private static string Tenant(GameId gameId, EnvironmentId environmentId) =>
$"{gameId.Value}/{environmentId.Value}";
private static IResult Error(RendezvousErrorCode code, int? retryAfterSeconds = null) => Results.Json(
new ApiError
{
Code = code,
Message = ErrorMessage(code),
RetryAfterSeconds = retryAfterSeconds,
},
ContractJson.Options,
statusCode: ErrorStatus(code));
@@ -392,6 +606,18 @@ internal static class ContractEndpoints
return Error(RendezvousErrorCode.AuthenticationRequired);
}
private static IResult RateLimited(HttpContext context)
{
int? retryAfterSeconds = int.TryParse(
context.Response.Headers.RetryAfter,
System.Globalization.NumberStyles.None,
System.Globalization.CultureInfo.InvariantCulture,
out int parsed)
? Math.Clamp(parsed, 1, 60)
: null;
return Error(RendezvousErrorCode.RateLimited, retryAfterSeconds);
}
private static int ErrorStatus(RendezvousErrorCode code) => code switch
{
RendezvousErrorCode.AuthenticationRequired => StatusCodes.Status401Unauthorized,
@@ -417,6 +643,7 @@ internal static class ContractEndpoints
RendezvousErrorCode.Expired => "The session lease has expired.",
RendezvousErrorCode.StaleHost => "The session has no fresh host presence.",
RendezvousErrorCode.IncompatibleProtocol => "The gameplay protocol is not enabled for this game.",
RendezvousErrorCode.RateLimited => "The request rate limit was exceeded.",
RendezvousErrorCode.CapacityExceeded => "The configured session capacity is currently exhausted.",
RendezvousErrorCode.ServiceUnavailable => "Session state is temporarily unavailable.",
RendezvousErrorCode.UnsupportedContractVersion => "The requested contract version is not supported.",
@@ -4,7 +4,8 @@ using Microsoft.AspNetCore.Diagnostics;
namespace FinalFactory.Rendezvous.Server.Http;
internal sealed class RendezvousExceptionHandler : IExceptionHandler
internal sealed class RendezvousExceptionHandler(
ILogger<RendezvousExceptionHandler> logger) : IExceptionHandler
{
public async ValueTask<bool> TryHandleAsync(
HttpContext httpContext,
@@ -17,21 +18,48 @@ internal sealed class RendezvousExceptionHandler : IExceptionHandler
}
bool invalidRequest = exception is BadHttpRequestException or JsonException;
httpContext.Response.StatusCode = invalidRequest
? StatusCodes.Status400BadRequest
: StatusCodes.Status500InternalServerError;
bool payloadTooLarge = exception is BadHttpRequestException
{
StatusCode: StatusCodes.Status413PayloadTooLarge,
};
httpContext.Response.StatusCode = payloadTooLarge
? StatusCodes.Status413PayloadTooLarge
: invalidRequest
? StatusCodes.Status400BadRequest
: StatusCodes.Status500InternalServerError;
LogRequestFailure(
logger,
payloadTooLarge ? "payload-too-large" : invalidRequest ? "invalid-request" : "internal-error",
httpContext.Response.StatusCode,
httpContext.Response.Headers["X-Rendezvous-Correlation-ID"].ToString() is { Length: > 0 } value
? value
: "unavailable");
await httpContext.Response.WriteAsJsonAsync(
new ApiError
{
Code = invalidRequest
? RendezvousErrorCode.InvalidRequest
: RendezvousErrorCode.InternalError,
Message = invalidRequest
? "The request body, route, or query value is invalid."
: "The service could not complete the request.",
Message = payloadTooLarge
? "The request body exceeds the supported size."
: invalidRequest
? "The request body, route, or query value is invalid."
: "The service could not complete the request.",
},
ContractJson.Options,
cancellationToken).ConfigureAwait(false);
return true;
}
private static readonly Action<ILogger, string, int, string, Exception?> RequestFailure =
LoggerMessage.Define<string, int, string>(
LogLevel.Warning,
new EventId(200, nameof(LogRequestFailure)),
"Request failed with {FailureKind} and HTTP status {StatusCode}; correlation {CorrelationId}");
private static void LogRequestFailure(
ILogger logger,
string failureKind,
int statusCode,
string correlationId) => RequestFailure(logger, failureKind, statusCode, correlationId, null);
}
@@ -0,0 +1,14 @@
using System.ComponentModel.DataAnnotations;
namespace FinalFactory.Rendezvous.Server.Observability;
internal sealed class AuditOptions
{
public const string SectionName = "Rendezvous:Audit";
[Range(100, 100_000)]
public int MaxEntries { get; set; } = 10_000;
[Range(1, 30)]
public int RetentionDays { get; set; } = 30;
}
@@ -0,0 +1,142 @@
using System.Security.Cryptography;
using System.Text;
using Microsoft.Extensions.Options;
namespace FinalFactory.Rendezvous.Server.Observability;
internal sealed class AuditTrail
{
private readonly object _gate = new();
private readonly LinkedList<AuditEntry> _entries = [];
private readonly AuditOptions _options;
private readonly TimeProvider _timeProvider;
private readonly ILogger<AuditTrail> _logger;
private readonly RendezvousTelemetry _telemetry;
public AuditTrail(
IOptions<AuditOptions> options,
ILogger<AuditTrail> logger,
RendezvousTelemetry telemetry,
TimeProvider? timeProvider = null)
{
_options = options.Value;
_logger = logger;
_telemetry = telemetry;
_timeProvider = timeProvider ?? TimeProvider.System;
}
public void Record(
string actorSubject,
string action,
string result,
string targetKind,
string targetIdentifier,
string correlationId)
{
DateTimeOffset now = _timeProvider.GetUtcNow();
AuditEntry entry = new(
now,
Fingerprint(actorSubject),
action,
result,
targetKind,
Fingerprint(targetIdentifier),
correlationId);
lock (_gate)
{
PurgeExpired(now);
while (_entries.Count >= _options.MaxEntries)
{
_entries.RemoveFirst();
}
_entries.AddLast(entry);
}
_telemetry.RecordAudit(action, result);
LogOperatorAction(
_logger,
entry.ActorFingerprint,
action,
result,
targetKind,
entry.TargetFingerprint,
correlationId);
}
public IReadOnlyDictionary<string, long> GetAggregateCounts()
{
lock (_gate)
{
PurgeExpired(_timeProvider.GetUtcNow());
return _entries
.GroupBy(static entry => $"{entry.Action}:{entry.Result}", StringComparer.Ordinal)
.ToDictionary(
static group => group.Key,
static group => (long)group.Count(),
StringComparer.Ordinal);
}
}
internal IReadOnlyList<AuditEntry> GetEntriesForTests()
{
lock (_gate)
{
PurgeExpired(_timeProvider.GetUtcNow());
return _entries.ToArray();
}
}
private void PurgeExpired(DateTimeOffset now)
{
DateTimeOffset oldest = now.AddDays(-_options.RetentionDays);
while (_entries.First is { Value.Timestamp: var timestamp }
&& timestamp < oldest)
{
_entries.RemoveFirst();
}
}
private static string Fingerprint(string value)
{
byte[] digest = SHA256.HashData(Encoding.UTF8.GetBytes(value));
return Convert.ToHexString(digest.AsSpan(0, 12));
}
private static readonly Action<ILogger, string, string, string, string, string, string, Exception?>
OperatorAction = LoggerMessage.Define<string, string, string, string, string, string>(
LogLevel.Information,
new EventId(100, nameof(LogOperatorAction)),
"Operator audit: actor {ActorFingerprint} action {Action} completed with {Result} for {TargetKind} target {TargetFingerprint}; correlation {CorrelationId}");
private static void LogOperatorAction(
ILogger logger,
string actorFingerprint,
string action,
string result,
string targetKind,
string targetFingerprint,
string correlationId) => OperatorAction(
logger,
actorFingerprint,
action,
result,
targetKind,
targetFingerprint,
correlationId,
null);
}
internal sealed record AuditEntry(
DateTimeOffset Timestamp,
string ActorFingerprint,
string Action,
string Result,
string TargetKind,
string TargetFingerprint,
string CorrelationId)
{
public override string ToString() =>
$"[AuditEntry {Action}/{Result}; actor and target fingerprinted]";
}
@@ -0,0 +1,30 @@
using FinalFactory.Rendezvous.Contracts;
namespace FinalFactory.Rendezvous.Server.Observability;
internal static class HealthEndpoints
{
public static IEndpointRouteBuilder MapRendezvousHealthEndpoints(
this IEndpointRouteBuilder endpoints)
{
endpoints.MapGet(
"/health/live",
static () => Results.Ok(new HealthResponse { Status = "live" }))
.Produces<HealthResponse>()
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.WithName("GetLiveness")
.WithTags("Health");
endpoints.MapGet(
"/health/ready",
static (RendezvousReadiness readiness) =>
!readiness.GetSnapshot().IsReady
? Results.StatusCode(StatusCodes.Status503ServiceUnavailable)
: Results.Ok(new HealthResponse { Status = "ready" }))
.Produces<HealthResponse>()
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces(StatusCodes.Status503ServiceUnavailable)
.WithName("GetReadiness")
.WithTags("Health");
return endpoints;
}
}
@@ -0,0 +1,41 @@
using FinalFactory.Rendezvous.Server.Provisioning;
using FinalFactory.Rendezvous.Server.State;
using FinalFactory.Rendezvous.Server.Transport;
using Microsoft.Extensions.Options;
namespace FinalFactory.Rendezvous.Server.Observability;
internal sealed class RendezvousReadiness(
UdpMediatorService mediator,
ProvisioningReadiness provisioning,
IEphemeralRendezvousStore state,
IOptions<UdpMediatorOptions> udpOptions)
{
public ReadinessSnapshot GetSnapshot()
{
bool ipv6Required = !string.IsNullOrWhiteSpace(udpOptions.Value.Ipv6ListenAddress);
return new ReadinessSnapshot(
HttpListenerReady: true,
UdpIpv4ListenerReady: mediator.LocalEndpoint is not null,
UdpIpv6ListenerReady: !ipv6Required || mediator.LocalIpv6Endpoint is not null,
ProvisioningReady: provisioning.IsReady,
StoreAvailable: state.IsAvailable,
Draining: state.IsDraining);
}
}
internal sealed record ReadinessSnapshot(
bool HttpListenerReady,
bool UdpIpv4ListenerReady,
bool UdpIpv6ListenerReady,
bool ProvisioningReady,
bool StoreAvailable,
bool Draining)
{
public bool IsReady => HttpListenerReady
&& UdpIpv4ListenerReady
&& UdpIpv6ListenerReady
&& ProvisioningReady
&& StoreAvailable
&& !Draining;
}
@@ -0,0 +1,126 @@
using System.Diagnostics;
using System.Diagnostics.Metrics;
using FinalFactory.Rendezvous.Server.State;
namespace FinalFactory.Rendezvous.Server.Observability;
internal sealed class RendezvousTelemetry : IDisposable
{
public const string MeterName = "FinalFactory.Rendezvous";
public const string ActivitySourceName = "FinalFactory.Rendezvous.Server";
private readonly InMemoryEphemeralRendezvousStore _store;
private readonly Meter _meter = new(MeterName, "1.0.0");
private readonly ActivitySource _activities = new(ActivitySourceName, "1.0.0");
private readonly Counter<long> _httpRequests;
private readonly Histogram<double> _httpDuration;
private readonly Counter<long> _udpResults;
private readonly Histogram<double> _udpDuration;
private readonly Counter<long> _limiterDrops;
private readonly Counter<long> _auditEvents;
private readonly Counter<long> _connectionOutcomes;
private readonly Counter<long> _operatorAuthentication;
private readonly Histogram<double> _pairingLatency;
public RendezvousTelemetry(InMemoryEphemeralRendezvousStore store)
{
_store = store;
_httpRequests = _meter.CreateCounter<long>("rendezvous.http.requests");
_httpDuration = _meter.CreateHistogram<double>(
"rendezvous.http.duration",
"ms");
_udpResults = _meter.CreateCounter<long>("rendezvous.udp.results");
_udpDuration = _meter.CreateHistogram<double>(
"rendezvous.udp.duration",
"ms");
_limiterDrops = _meter.CreateCounter<long>("rendezvous.limiter.drops");
_auditEvents = _meter.CreateCounter<long>("rendezvous.audit.events");
_connectionOutcomes = _meter.CreateCounter<long>("rendezvous.connection.outcomes");
_operatorAuthentication = _meter.CreateCounter<long>("rendezvous.operator.authentication");
_pairingLatency = _meter.CreateHistogram<double>(
"rendezvous.pairing.latency",
"ms");
_meter.CreateObservableGauge(
"rendezvous.store.active_listings",
() => _store.GetMetricsSnapshot().ActiveListings);
_meter.CreateObservableGauge(
"rendezvous.store.active_leases",
() => _store.GetMetricsSnapshot().ActiveListings);
_meter.CreateObservableGauge(
"rendezvous.store.active_attempts",
() => _store.GetMetricsSnapshot().ActiveJoinAttempts);
_meter.CreateObservableGauge(
"rendezvous.queue.depth",
() => _store.GetMetricsSnapshot().ActiveJoinAttempts);
_meter.CreateObservableGauge(
"rendezvous.store.replay_markers",
() => _store.GetMetricsSnapshot().ReplayMarkers);
_meter.CreateObservableGauge(
"rendezvous.store.available",
() => _store.GetMetricsSnapshot().IsAvailable ? 1 : 0);
_meter.CreateObservableCounter(
"rendezvous.store.expiry_churn",
() => _store.GetMetricsSnapshot().ExpiryChurn);
}
public Activity? StartActivity(string name, ActivityKind kind = ActivityKind.Internal) =>
_activities.StartActivity(name, kind);
public void RecordHttp(string operation, int statusCode, double elapsedMilliseconds)
{
TagList tags = new()
{
{ "operation", operation },
{ "status_code", statusCode },
};
_httpRequests.Add(1, tags);
_httpDuration.Record(elapsedMilliseconds, tags);
}
public void RecordUdp(string operation, string result, double elapsedMilliseconds)
{
TagList tags = new()
{
{ "operation", operation },
{ "result", result },
};
_udpResults.Add(1, tags);
_udpDuration.Record(elapsedMilliseconds, tags);
}
public void RecordLimiterDrop(string transport, string partition) =>
_limiterDrops.Add(1, new TagList
{
{ "transport", transport },
{ "partition", partition },
});
public void RecordAudit(string action, string result) =>
_auditEvents.Add(1, new TagList
{
{ "action", action },
{ "result", result },
});
public void RecordConnectionOutcome(string outcome, string elapsedBucket) =>
_connectionOutcomes.Add(1, new TagList
{
{ "outcome", outcome },
{ "elapsed_bucket", elapsedBucket },
});
public void RecordOperatorAuthentication(string result) =>
_operatorAuthentication.Add(1, new TagList
{
{ "result", result },
});
public void RecordPairingLatency(double elapsedMilliseconds) =>
_pairingLatency.Record(elapsedMilliseconds);
public void Dispose()
{
_activities.Dispose();
_meter.Dispose();
}
}
@@ -0,0 +1,32 @@
using System.Diagnostics;
namespace FinalFactory.Rendezvous.Server.Observability;
internal sealed class TelemetryMiddleware(
RequestDelegate next,
RendezvousTelemetry telemetry)
{
public async Task InvokeAsync(HttpContext context)
{
string operation = context.GetEndpoint()?.Metadata.GetMetadata<IEndpointNameMetadata>()
?.EndpointName ?? "Unmatched";
long started = Stopwatch.GetTimestamp();
using Activity? activity = telemetry.StartActivity(
$"HTTP {operation}",
ActivityKind.Server);
string correlationId = activity?.TraceId.ToString() ?? Guid.NewGuid().ToString("N");
context.Response.Headers["X-Rendezvous-Correlation-ID"] = correlationId;
activity?.SetTag("rendezvous.operation", operation);
try
{
await next(context).ConfigureAwait(false);
}
finally
{
telemetry.RecordHttp(
operation,
context.Response.StatusCode,
Stopwatch.GetElapsedTime(started).TotalMilliseconds);
}
}
}
@@ -0,0 +1,428 @@
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.Provisioning;
using FinalFactory.Rendezvous.Server.State;
using Microsoft.AspNetCore.Mvc;
namespace FinalFactory.Rendezvous.Server.Operations;
internal static class OperatorEndpoints
{
private const string CorrelationHeader = "X-Rendezvous-Correlation-ID";
public static IEndpointRouteBuilder MapOperatorEndpoints(this IEndpointRouteBuilder endpoints)
{
RouteGroupBuilder group = endpoints.MapGroup("/v1/operator").WithTags("Operator");
group.MapGet("/status", GetStatus)
.Produces<OperatorStatusResponse>()
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.WithName("GetOperatorStatus");
group.MapPost("/listings/revoke", RevokeListing)
.Accepts<RevokeListingRequest>("application/json")
.Produces<OperatorActionResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("RevokeOperatorListing");
group.MapPost("/principals/revoke", RevokePrincipal)
.Accepts<RevokePrincipalRequest>("application/json")
.Produces<OperatorActionResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.Produces<ApiError>(StatusCodes.Status503ServiceUnavailable)
.WithName("RevokeOperatorPrincipal");
group.MapPost("/keys/revoke", RevokeSigningKey)
.Accepts<RevokeSigningKeyRequest>("application/json")
.Produces<OperatorActionResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.WithName("RevokeOperatorSigningKey");
group.MapPost("/drain", BeginDrain)
.Accepts<BeginDrainRequest>("application/json")
.Produces<OperatorActionResponse>()
.Produces<ApiError>(StatusCodes.Status400BadRequest)
.Produces<ApiError>(StatusCodes.Status413PayloadTooLarge)
.Produces<ApiError>(StatusCodes.Status401Unauthorized)
.Produces<ApiError>(StatusCodes.Status403Forbidden)
.Produces<ApiError>(StatusCodes.Status404NotFound)
.Produces<ApiError>(StatusCodes.Status429TooManyRequests)
.WithName("BeginOperatorDrain");
return endpoints;
}
private static IResult GetStatus(
[FromHeader(Name = "Authorization")] string? authorization,
[FromServices] PrincipalCredentialService credentials,
[FromServices] IWallClock clock,
[FromServices] OperatorService service,
[FromServices] AuditTrail audit,
[FromServices] RendezvousTelemetry telemetry,
HttpContext context)
{
if (!TryAuthorize(
authorization,
OperatorPermission.ReadPolicy,
"inspect-status",
credentials,
clock,
audit,
telemetry,
context,
out OperatorPrincipal? principal,
out IResult? failure))
{
return failure!;
}
OperatorStatusResponse response = service.GetStatus();
audit.Record(
principal!.Subject,
"inspect-status",
"succeeded",
"service",
"rendezvous",
Correlation(context));
return Results.Ok(response);
}
private static IResult RevokeListing(
[FromBody] RevokeListingRequest request,
[FromHeader(Name = "Authorization")] string? authorization,
[FromServices] PrincipalCredentialService credentials,
[FromServices] IWallClock clock,
[FromServices] OperatorService service,
[FromServices] AuditTrail audit,
[FromServices] RendezvousTelemetry telemetry,
HttpContext context,
CancellationToken cancellationToken)
{
if (!TryAuthorize(
authorization,
OperatorPermission.RevokePublisher,
"revoke-listing",
credentials,
clock,
audit,
telemetry,
context,
out OperatorPrincipal? principal,
out IResult? failure))
{
return failure!;
}
bool valid = SessionListingId.TryParse(request.ListingId, out SessionListingId listingId);
if (!valid || !string.Equals(request.ListingId, request.ConfirmListingId, StringComparison.Ordinal))
{
AuditRejected(audit, principal!, "revoke-listing", "listing", request.ListingId, context);
return BadRequest("A valid listing ID and an exact repeated confirmation are required.");
}
StoreResult<bool> result = service.RevokeListing(listingId, cancellationToken);
return StoreActionResult(
result.Code,
audit,
principal!,
"revoke-listing",
"listing",
request.ListingId,
context,
affectedResources: result.Succeeded ? 1 : null);
}
private static IResult RevokePrincipal(
[FromBody] RevokePrincipalRequest request,
[FromHeader(Name = "Authorization")] string? authorization,
[FromServices] PrincipalCredentialService credentials,
[FromServices] IWallClock clock,
[FromServices] OperatorService service,
[FromServices] AuditTrail audit,
[FromServices] RendezvousTelemetry telemetry,
HttpContext context,
CancellationToken cancellationToken)
{
if (!TryAuthorize(
authorization,
OperatorPermission.RevokePublisher,
"revoke-principal",
credentials,
clock,
audit,
telemetry,
context,
out OperatorPrincipal? principal,
out IResult? failure))
{
return failure!;
}
bool safeSubject = request.Subject is { Length: > 0 and <= 128 }
&& request.Subject.All(static character => character is >= '!' and <= '~');
if (!safeSubject
|| !string.Equals(request.Subject, request.ConfirmSubject, StringComparison.Ordinal)
|| request.LifetimeSeconds is < 1 or > 600)
{
AuditRejected(audit, principal!, "revoke-principal", "principal", request.Subject, context);
return BadRequest("A valid subject, exact repeated confirmation, and 1-600 second lifetime are required.");
}
StoreResult<int> result = service.RevokePrincipal(
request.Subject,
TimeSpan.FromSeconds(request.LifetimeSeconds),
cancellationToken);
return StoreActionResult(
result.Code,
audit,
principal!,
"revoke-principal",
"principal",
request.Subject,
context,
result.Value);
}
private static IResult RevokeSigningKey(
[FromBody] RevokeSigningKeyRequest request,
[FromHeader(Name = "Authorization")] string? authorization,
[FromServices] PrincipalCredentialService credentials,
[FromServices] IWallClock clock,
[FromServices] OperatorService service,
[FromServices] AuditTrail audit,
[FromServices] RendezvousTelemetry telemetry,
HttpContext context)
{
if (!TryAuthorize(
authorization,
OperatorPermission.RotateKeys,
"revoke-signing-key",
credentials,
clock,
audit,
telemetry,
context,
out OperatorPrincipal? principal,
out IResult? failure))
{
return failure!;
}
bool safeKeyId = request.KeyId is { Length: > 0 and <= 64 }
&& request.KeyId.All(static character => character is
>= 'A' and <= 'Z'
or >= 'a' and <= 'z'
or >= '0' and <= '9'
or '-'
or '_');
if (!safeKeyId || !string.Equals(request.KeyId, request.ConfirmKeyId, StringComparison.Ordinal))
{
AuditRejected(audit, principal!, "revoke-signing-key", "signing-key", request.KeyId, context);
return BadRequest("A valid key ID and an exact repeated confirmation are required.");
}
bool revoked = service.RevokeSigningKey(request.KeyId);
string result = revoked ? "succeeded" : "not-found";
audit.Record(
principal!.Subject,
"revoke-signing-key",
result,
"signing-key",
request.KeyId,
Correlation(context));
return revoked
? Results.Ok(new OperatorActionResponse { Status = "completed" })
: Error(RendezvousErrorCode.NotFound, "The requested resource was not found.");
}
private static IResult BeginDrain(
[FromBody] BeginDrainRequest request,
[FromHeader(Name = "Authorization")] string? authorization,
[FromServices] PrincipalCredentialService credentials,
[FromServices] IWallClock clock,
[FromServices] OperatorService service,
[FromServices] AuditTrail audit,
[FromServices] RendezvousTelemetry telemetry,
HttpContext context,
CancellationToken cancellationToken)
{
if (!TryAuthorize(
authorization,
OperatorPermission.ManagePolicy,
"begin-drain",
credentials,
clock,
audit,
telemetry,
context,
out OperatorPrincipal? principal,
out IResult? failure))
{
return failure!;
}
if (!string.Equals(request.Confirmation, "DRAIN", StringComparison.Ordinal))
{
AuditRejected(audit, principal!, "begin-drain", "service", "rendezvous", context);
return BadRequest("The confirmation value must be exactly 'DRAIN'.");
}
service.BeginDrain(cancellationToken);
audit.Record(
principal!.Subject,
"begin-drain",
"succeeded",
"service",
"rendezvous",
Correlation(context));
return Results.Ok(new OperatorActionResponse { Status = "draining" });
}
private static bool TryAuthorize(
string? authorization,
OperatorPermission requiredPermission,
string operation,
PrincipalCredentialService credentials,
IWallClock clock,
AuditTrail audit,
RendezvousTelemetry telemetry,
HttpContext context,
out OperatorPrincipal? principal,
out IResult? failure)
{
principal = null;
failure = null;
const string prefix = "Bearer ";
if (authorization is null
|| !authorization.StartsWith(prefix, StringComparison.OrdinalIgnoreCase))
{
telemetry.RecordOperatorAuthentication("rejected");
failure = AuthenticationRequired(context);
return false;
}
CredentialValidationResult validation = credentials.Validate(
authorization[prefix.Length..],
clock.UtcNow);
if (!validation.IsValid || validation.Principal is not OperatorPrincipal candidate)
{
telemetry.RecordOperatorAuthentication("rejected");
failure = AuthenticationRequired(context);
return false;
}
if (!candidate.Permissions.Contains(requiredPermission))
{
telemetry.RecordOperatorAuthentication("forbidden");
audit.Record(
candidate.Subject,
operation,
"forbidden",
"operator-operation",
operation,
Correlation(context));
failure = Error(RendezvousErrorCode.Forbidden, "The operator is not authorized for this operation.");
return false;
}
telemetry.RecordOperatorAuthentication("accepted");
principal = candidate;
return true;
}
private static IResult StoreActionResult(
StoreResultCode code,
AuditTrail audit,
OperatorPrincipal principal,
string action,
string targetKind,
string targetIdentifier,
HttpContext context,
int? affectedResources)
{
string auditResult = code == StoreResultCode.Success
? "succeeded"
: code.ToString().ToLowerInvariant();
audit.Record(
principal.Subject,
action,
auditResult,
targetKind,
targetIdentifier,
Correlation(context));
return code switch
{
StoreResultCode.Success => Results.Ok(new OperatorActionResponse
{
Status = "completed",
AffectedResources = affectedResources,
}),
StoreResultCode.NotFound => Error(
RendezvousErrorCode.NotFound,
"The requested resource was not found."),
StoreResultCode.CapacityExceeded => Error(
RendezvousErrorCode.CapacityExceeded,
"The operation could not be retained within the configured capacity."),
StoreResultCode.ServiceUnavailable or StoreResultCode.Draining => Error(
RendezvousErrorCode.ServiceUnavailable,
"The service is not available for this operation."),
_ => Error(RendezvousErrorCode.Conflict, "The operation could not be completed."),
};
}
private static void AuditRejected(
AuditTrail audit,
OperatorPrincipal principal,
string action,
string targetKind,
string? targetIdentifier,
HttpContext context) => audit.Record(
principal.Subject,
action,
"rejected",
targetKind,
targetIdentifier ?? string.Empty,
Correlation(context));
private static string Correlation(HttpContext context) =>
context.Response.Headers[CorrelationHeader].ToString() is { Length: > 0 } value
? value
: "unavailable";
private static IResult AuthenticationRequired(HttpContext context)
{
context.Response.Headers.WWWAuthenticate = "Bearer realm=\"operator\"";
return Error(
RendezvousErrorCode.AuthenticationRequired,
"A valid operator bearer credential is required.");
}
private static IResult BadRequest(string message) => Error(RendezvousErrorCode.InvalidRequest, message);
private static IResult Error(RendezvousErrorCode code, string message) => Results.Json(
new ApiError { Code = code, Message = message },
ContractJson.Options,
statusCode: code switch
{
RendezvousErrorCode.AuthenticationRequired => StatusCodes.Status401Unauthorized,
RendezvousErrorCode.Forbidden => StatusCodes.Status403Forbidden,
RendezvousErrorCode.NotFound => StatusCodes.Status404NotFound,
RendezvousErrorCode.CapacityExceeded => StatusCodes.Status429TooManyRequests,
RendezvousErrorCode.ServiceUnavailable => StatusCodes.Status503ServiceUnavailable,
RendezvousErrorCode.Conflict => StatusCodes.Status409Conflict,
_ => StatusCodes.Status400BadRequest,
});
}
@@ -0,0 +1,95 @@
namespace FinalFactory.Rendezvous.Server.Operations;
internal sealed record OperatorStatusResponse
{
public required string Status { get; init; }
public required OperatorCompatibilityResponse Compatibility { get; init; }
public required OperatorReadinessResponse Readiness { get; init; }
public required OperatorStoreResponse Store { get; init; }
public required IReadOnlyList<OperatorTenantResponse> Tenants { get; init; }
public required IReadOnlyList<OperatorSigningKeyResponse> SigningKeys { get; init; }
public required IReadOnlyDictionary<string, long> AuditCounts { get; init; }
}
internal sealed record OperatorCompatibilityResponse
{
public required string ServerVersion { get; init; }
public required string MinimumClientVersion { get; init; }
public required int MaximumClientMajorVersion { get; init; }
public required IReadOnlyList<int> HttpContractVersions { get; init; }
public required IReadOnlyList<int> UdpContractVersions { get; init; }
public required IReadOnlyList<int> ConnectionTicketFormatVersions { get; init; }
public required int LiteNetLibMajorVersion { get; init; }
public required string GameplayProtocolCompatibility { get; init; }
}
internal sealed record OperatorReadinessResponse
{
public required bool HttpListener { get; init; }
public required bool UdpIpv4Listener { get; init; }
public required bool UdpIpv6Listener { get; init; }
public required bool Provisioning { get; init; }
public required bool Store { get; init; }
public required bool Draining { get; init; }
}
internal sealed record OperatorStoreResponse
{
public required int ActiveListings { get; init; }
public required int FreshPresenceBindings { get; init; }
public required int ActiveJoinAttempts { get; init; }
public required int RetainedOutcomeReports { get; init; }
public required int ReplayMarkers { get; init; }
public required int PrincipalRevocations { get; init; }
public required int IdempotencyEntries { get; init; }
public required long MaintenanceSweeps { get; init; }
public required long ExpiryChurn { get; init; }
}
internal sealed record OperatorTenantResponse
{
public required string GameId { get; init; }
public required string EnvironmentId { get; init; }
public required string Status { get; init; }
}
internal sealed record OperatorSigningKeyResponse
{
public required string KeyId { get; init; }
public required string Status { get; init; }
public required DateTimeOffset SignUntil { get; init; }
public required DateTimeOffset VerifyUntil { get; init; }
public string? GameId { get; init; }
public string? EnvironmentId { get; init; }
public required IReadOnlyList<string> CredentialKinds { get; init; }
}
internal sealed record OperatorActionResponse
{
public required string Status { get; init; }
public int? AffectedResources { get; init; }
}
internal sealed record RevokeListingRequest
{
public required string ListingId { get; init; }
public required string ConfirmListingId { get; init; }
}
internal sealed record RevokePrincipalRequest
{
public required string Subject { get; init; }
public required string ConfirmSubject { get; init; }
public required int LifetimeSeconds { get; init; }
}
internal sealed record RevokeSigningKeyRequest
{
public required string KeyId { get; init; }
public required string ConfirmKeyId { get; init; }
}
internal sealed record BeginDrainRequest
{
public required string Confirmation { get; init; }
}
@@ -0,0 +1,82 @@
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.Provisioning;
using FinalFactory.Rendezvous.Server.State;
namespace FinalFactory.Rendezvous.Server.Operations;
internal sealed class OperatorService(
InMemoryEphemeralRendezvousStore store,
ProvisioningRuntime provisioning,
RendezvousReadiness readiness,
AuditTrail audit,
IWallClock clock)
{
public OperatorStatusResponse GetStatus()
{
ReadinessSnapshot readinessSnapshot = readiness.GetSnapshot();
EphemeralStoreSnapshot storeSnapshot = store.GetSnapshot();
return new OperatorStatusResponse
{
Status = readinessSnapshot.IsReady ? "ready" : "not-ready",
Compatibility = ReleaseCompatibility.CreateResponse(),
Readiness = new OperatorReadinessResponse
{
HttpListener = readinessSnapshot.HttpListenerReady,
UdpIpv4Listener = readinessSnapshot.UdpIpv4ListenerReady,
UdpIpv6Listener = readinessSnapshot.UdpIpv6ListenerReady,
Provisioning = readinessSnapshot.ProvisioningReady,
Store = readinessSnapshot.StoreAvailable,
Draining = readinessSnapshot.Draining,
},
Store = new OperatorStoreResponse
{
ActiveListings = storeSnapshot.ActiveListings,
FreshPresenceBindings = storeSnapshot.FreshPresenceBindings,
ActiveJoinAttempts = storeSnapshot.ActiveJoinAttempts,
RetainedOutcomeReports = storeSnapshot.RetainedOutcomeReports,
ReplayMarkers = storeSnapshot.ReplayMarkers,
PrincipalRevocations = storeSnapshot.PrincipalRevocations,
IdempotencyEntries = storeSnapshot.IdempotencyEntries,
MaintenanceSweeps = storeSnapshot.MaintenanceSweeps,
ExpiryChurn = storeSnapshot.ExpiryChurn,
},
Tenants = provisioning.Policies.EnabledPolicies
.OrderBy(static policy => policy.GameId.Value, StringComparer.Ordinal)
.ThenBy(static policy => policy.EnvironmentId.Value, StringComparer.Ordinal)
.Select(static policy => new OperatorTenantResponse
{
GameId = policy.GameId.Value,
EnvironmentId = policy.EnvironmentId.Value,
Status = "enabled",
})
.ToArray(),
SigningKeys = provisioning.SigningKeys.GetStatuses(clock.UtcNow)
.Select(static key => new OperatorSigningKeyResponse
{
KeyId = key.KeyId,
Status = key.Status,
SignUntil = key.SignUntil,
VerifyUntil = key.VerifyUntil,
GameId = key.GameId,
EnvironmentId = key.EnvironmentId,
CredentialKinds = key.CredentialKinds,
})
.ToArray(),
AuditCounts = audit.GetAggregateCounts(),
};
}
public StoreResult<bool> RevokeListing(
SessionListingId listingId,
CancellationToken cancellationToken) => store.RevokeListing(listingId, cancellationToken);
public StoreResult<int> RevokePrincipal(
string subject,
TimeSpan lifetime,
CancellationToken cancellationToken) => store.RevokePrincipal(subject, lifetime, cancellationToken);
public bool RevokeSigningKey(string keyId) => provisioning.SigningKeys.Revoke(keyId);
public void BeginDrain(CancellationToken cancellationToken) => store.BeginDrain(cancellationToken);
}
@@ -0,0 +1,41 @@
using System.Reflection;
using FinalFactory.Rendezvous.Contracts;
namespace FinalFactory.Rendezvous.Server.Operations;
internal static class ReleaseCompatibility
{
private static readonly Assembly ServerAssembly = typeof(ReleaseCompatibility).Assembly;
internal static string MinimumClientVersion => Metadata("RendezvousMinimumClientVersion");
internal static int MaximumClientMajorVersion => MetadataInteger("RendezvousMaximumClientMajorVersion");
internal static int UdpContractVersion => MetadataInteger("RendezvousUdpContractVersion");
internal static int ConnectionTicketFormatVersion => MetadataInteger("RendezvousConnectionTicketFormatVersion");
internal static int LiteNetLibMajorVersion => MetadataInteger("RendezvousLiteNetLibMajorVersion");
internal static OperatorCompatibilityResponse CreateResponse() => new()
{
ServerVersion = ServerAssembly
.GetCustomAttribute<AssemblyInformationalVersionAttribute>()?
.InformationalVersion.Split('+', 2)[0]
?? ServerAssembly.GetName().Version?.ToString(3)
?? "unknown",
MinimumClientVersion = MinimumClientVersion,
MaximumClientMajorVersion = MaximumClientMajorVersion,
HttpContractVersions = [ContractLimits.ContractVersion],
UdpContractVersions = [UdpContractVersion],
ConnectionTicketFormatVersions = [ConnectionTicketFormatVersion],
LiteNetLibMajorVersion = LiteNetLibMajorVersion,
GameplayProtocolCompatibility = "exact-per-tenant",
};
private static string Metadata(string key) => ServerAssembly
.GetCustomAttributes<AssemblyMetadataAttribute>()
.Single(attribute => string.Equals(attribute.Key, key, StringComparison.Ordinal))
.Value
?? throw new InvalidOperationException($"Assembly metadata {key} has no value.");
private static int MetadataInteger(string key) => int.Parse(
Metadata(key),
System.Globalization.CultureInfo.InvariantCulture);
}
+167 -24
View File
@@ -1,16 +1,24 @@
using System.Net;
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Abuse;
using FinalFactory.Rendezvous.Server.Browser;
using FinalFactory.Rendezvous.Server.ConnectionOutcomes;
using FinalFactory.Rendezvous.Server.Deployment;
using FinalFactory.Rendezvous.Server.Http;
using FinalFactory.Rendezvous.Server.JoinAttempts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.Operations;
using FinalFactory.Rendezvous.Server.Provisioning;
using FinalFactory.Rendezvous.Server.Sessions;
using FinalFactory.Rendezvous.Server.State;
using FinalFactory.Rendezvous.Server.Transport;
using Microsoft.AspNetCore.HttpOverrides;
using Microsoft.OpenApi;
WebApplicationBuilder builder = WebApplication.CreateBuilder(args);
builder.Logging.AddFilter(
"Microsoft.AspNetCore.Diagnostics.ExceptionHandlerMiddleware",
LogLevel.None);
bool isOpenApiGeneration = string.Equals(
System.Reflection.Assembly.GetEntryAssembly()?.GetName().Name,
"GetDocument.Insider",
@@ -59,6 +67,14 @@ builder.Services.AddOpenApi("v1", static options =>
In = ParameterLocation.Header,
Description = "Attempt-scoped client capability returned only to the joining caller.",
};
const string operatorSchemeName = "OperatorBearer";
document.Components.SecuritySchemes[operatorSchemeName] = new OpenApiSecurityScheme
{
Type = SecuritySchemeType.Http,
Scheme = "bearer",
BearerFormat = "rv1 operator credential",
Description = "Operator-only credential with an explicit permission set.",
};
HashSet<string> securedOperations = new(StringComparer.Ordinal)
{
@@ -67,8 +83,17 @@ builder.Services.AddOpenApi("v1", static options =>
"UpdateSession",
"DeleteSession",
};
HashSet<string> operatorOperations = new(StringComparer.Ordinal)
{
"GetOperatorStatus",
"RevokeOperatorListing",
"RevokeOperatorPrincipal",
"RevokeOperatorSigningKey",
"BeginOperatorDrain",
};
OpenApiSecuritySchemeReference reference = new(schemeName, document, null);
OpenApiSecuritySchemeReference attemptReference = new(attemptSchemeName, document, null);
OpenApiSecuritySchemeReference operatorReference = new(operatorSchemeName, document, null);
foreach (OpenApiPathItem path in document.Paths.Values)
{
if (path.Operations is null)
@@ -96,6 +121,60 @@ builder.Services.AddOpenApi("v1", static options =>
[attemptReference] = [],
});
}
foreach (OpenApiOperation operation in path.Operations.Values.Where(
operation => operatorOperations.Contains(
operation.OperationId ?? string.Empty)))
{
operation.Security ??= [];
operation.Security.Add(new OpenApiSecurityRequirement
{
[operatorReference] = [],
});
}
foreach (OpenApiOperation operation in path.Operations.Values)
{
if (operation.Responses is null)
{
continue;
}
foreach ((string status, IOpenApiResponse response) in operation.Responses)
{
if (response is not OpenApiResponse concreteResponse)
{
continue;
}
concreteResponse.Headers ??=
new Dictionary<string, IOpenApiHeader>(StringComparer.OrdinalIgnoreCase);
concreteResponse.Headers["X-Rendezvous-Correlation-ID"] = new OpenApiHeader
{
Description = "Safe request correlation identifier generated by the service.",
Schema = new OpenApiSchema
{
Type = JsonSchemaType.String,
},
};
if (string.Equals(
status,
StatusCodes.Status429TooManyRequests.ToString(
System.Globalization.CultureInfo.InvariantCulture),
StringComparison.Ordinal))
{
concreteResponse.Headers["Retry-After"] = new OpenApiHeader
{
Description = "Whole seconds before the caller should retry (1-60).",
Schema = new OpenApiSchema
{
Type = JsonSchemaType.Integer,
Format = "int32",
},
};
}
}
}
}
return Task.CompletedTask;
@@ -107,15 +186,90 @@ builder.Services.Configure<RouteHandlerOptions>(static options =>
options.ThrowOnBadRequest = true);
builder.Services.AddProblemDetails();
builder.Services.AddExceptionHandler<RendezvousExceptionHandler>();
builder.WebHost.ConfigureKestrel(static options =>
options.Limits.MaxRequestBodySize = ContractLimits.HttpRequestMaxBytes);
builder.Services
.AddOptions<AbuseProtectionOptions>()
.BindConfiguration(AbuseProtectionOptions.SectionName)
.ValidateDataAnnotations()
.Validate(
options => options.HttpOptionalRequestsPerWindow
< options.HttpGlobalRequestsPerWindow,
"The optional HTTP request budget must leave global capacity for lease operations.")
.Validate(
options => options.HttpOptionalConcurrency < options.HttpGlobalConcurrency,
"The optional HTTP concurrency budget must leave global capacity for lease operations.")
.Validate(
options => options.HttpOptionalIpPrefixRequestsPerWindow
< options.HttpIpPrefixRequestsPerWindow,
"The optional HTTP source budget must leave capacity for lease operations.")
.Validate(
options => options.HttpOptionalIpPrefixConcurrency
< options.HttpIpPrefixConcurrency,
"The optional HTTP source concurrency must leave capacity for lease operations.")
.Validate(
options => options.CriticalTrackedKeyReserve >= 16
&& options.UdpTrackedKeyLimit + options.CriticalTrackedKeyReserve
< options.MaxTrackedKeys,
"The tracked-key reserve must leave at least 16 keys for critical operations.")
.Validate(
options => options.TrustedProxyAddresses is { Length: <= 32 } addresses
&& addresses.All(
static value => IPAddress.TryParse(value, out _)),
"Trusted proxy addresses must contain at most 32 literal IP addresses.")
.Validate(
options => options.OperatorAllowedAddresses is { Length: <= 32 } addresses
&& addresses.All(
static value => IPAddress.TryParse(value, out _)),
"Operator allowed addresses must contain at most 32 literal IP addresses.")
.ValidateOnStart();
builder.Services.AddSingleton<AbuseProtectionService>();
builder.Services
.AddOptions<AuditOptions>()
.BindConfiguration(AuditOptions.SectionName)
.ValidateDataAnnotations()
.ValidateOnStart();
AbuseProtectionOptions configuredAbuseProtection = builder.Configuration
.GetSection(AbuseProtectionOptions.SectionName)
.Get<AbuseProtectionOptions>() ?? new AbuseProtectionOptions();
builder.Services.Configure<ForwardedHeadersOptions>(options =>
TrustedProxyForwarding.Configure(options, configuredAbuseProtection));
DeploymentOptions deploymentOptions = builder.Configuration
.GetSection(DeploymentOptions.SectionName)
.Get<DeploymentOptions>() ?? new DeploymentOptions();
if (!builder.Environment.IsDevelopment() && !isOpenApiGeneration)
{
IReadOnlyList<string> deploymentErrors = deploymentOptions.ValidateProduction(
configuredAbuseProtection,
builder.Configuration["AllowedHosts"]);
if (deploymentErrors.Count > 0)
{
throw new DeploymentConfigurationException(deploymentErrors);
}
}
builder.Services.AddSingleton(Microsoft.Extensions.Options.Options.Create(deploymentOptions));
builder.Services.Configure<HostOptions>(options =>
options.ShutdownTimeout = TimeSpan.FromSeconds(deploymentOptions.DrainDeadlineSeconds + 10));
SystemRendezvousClock rendezvousClock = new();
EphemeralStoreOptions stateOptions = new();
EphemeralStoreOptions stateOptions = new()
{
GracefulDrainLifetime = TimeSpan.FromSeconds(deploymentOptions.DrainDeadlineSeconds),
};
InMemoryEphemeralRendezvousStore stateStore = new(
stateOptions,
rendezvousClock,
rendezvousClock);
builder.Services.AddSingleton(stateStore);
builder.Services.AddSingleton<IEphemeralRendezvousStore>(stateStore);
builder.Services.AddSingleton<IWallClock>(rendezvousClock);
builder.Services.AddSingleton<IMonotonicClock>(rendezvousClock);
builder.Services.AddSingleton<RendezvousTelemetry>();
builder.Services.AddSingleton<AuditTrail>();
builder.Services.AddSingleton<RendezvousReadiness>();
if (isOpenApiGeneration)
{
@@ -148,6 +302,7 @@ else
builder.Services.AddSingleton<JoinAttemptService>();
builder.Services.AddSingleton<ConnectionOutcomeMetrics>();
builder.Services.AddSingleton<ConnectionOutcomeService>();
builder.Services.AddSingleton<OperatorService>();
builder.Services.AddSingleton(new ProvisioningReadiness(true));
}
@@ -171,36 +326,24 @@ if (!isOpenApiGeneration)
builder.Services.AddSingleton<NatMediationProcessor>();
builder.Services.AddHostedService(static services =>
services.GetRequiredService<UdpMediatorService>());
// Hosted services stop in reverse registration order. Drain must complete while
// Kestrel and the UDP mediator are still able to finish bounded in-flight work.
builder.Services.AddHostedService<GracefulDrainService>();
}
WebApplication app = builder.Build();
app.Lifetime.ApplicationStopping.Register(() => stateStore.BeginDrain());
if (TrustedProxyForwarding.IsEnabled(configuredAbuseProtection))
{
app.UseForwardedHeaders();
}
app.UseMiddleware<TelemetryMiddleware>();
app.UseExceptionHandler();
app.UseMiddleware<HttpAbuseProtectionMiddleware>();
app.MapOpenApi();
app.MapRendezvousContractEndpoints();
app.MapGet(
"/health/live",
static () => Results.Ok(new HealthResponse { Status = "live" }))
.Produces<HealthResponse>()
.WithName("GetLiveness")
.WithTags("Health");
app.MapGet(
"/health/ready",
static (
UdpMediatorService mediator,
ProvisioningReadiness provisioning,
IEphemeralRendezvousStore state) =>
mediator.LocalEndpoint is null
|| !provisioning.IsReady
|| !state.IsAvailable
|| state.IsDraining
? Results.StatusCode(StatusCodes.Status503ServiceUnavailable)
: Results.Ok(new HealthResponse { Status = "ready" }))
.Produces<HealthResponse>()
.Produces(StatusCodes.Status503ServiceUnavailable)
.WithName("GetReadiness")
.WithTags("Health");
app.MapOperatorEndpoints();
app.MapRendezvousHealthEndpoints();
await app.RunAsync();
@@ -1,3 +1,4 @@
using System.Runtime.CompilerServices;
[assembly: InternalsVisibleTo("FinalFactory.Rendezvous.Tests")]
[assembly: InternalsVisibleTo("FinalFactory.Rendezvous.Capacity")]
@@ -40,18 +40,32 @@ internal sealed class SecretMaterial : IDisposable
internal sealed class EnvironmentSecretProvider : ISecretProvider
{
private const string Prefix = "env:";
private const string EnvironmentPrefix = "env:";
private const string FilePrefix = "file:";
private const int MaximumSecretBytes = 4096;
public bool TryGetSecret(string reference, out SecretMaterial? secret)
{
secret = null;
if (!reference.StartsWith(Prefix, StringComparison.Ordinal)
|| reference.Length == Prefix.Length)
if (reference.StartsWith(EnvironmentPrefix, StringComparison.Ordinal)
&& reference.Length > EnvironmentPrefix.Length)
{
return false;
return TryGetEnvironmentSecret(reference[EnvironmentPrefix.Length..], out secret);
}
string? encoded = Environment.GetEnvironmentVariable(reference[Prefix.Length..]);
if (reference.StartsWith(FilePrefix, StringComparison.Ordinal)
&& reference.Length > FilePrefix.Length)
{
return TryGetFileSecret(reference[FilePrefix.Length..], out secret);
}
return false;
}
private static bool TryGetEnvironmentSecret(string variableName, out SecretMaterial? secret)
{
secret = null;
string? encoded = Environment.GetEnvironmentVariable(variableName);
if (string.IsNullOrEmpty(encoded))
{
return false;
@@ -60,6 +74,12 @@ internal sealed class EnvironmentSecretProvider : ISecretProvider
try
{
byte[] bytes = Convert.FromBase64String(encoded);
if (bytes.Length is 0 or > MaximumSecretBytes)
{
CryptographicOperations.ZeroMemory(bytes);
return false;
}
secret = new SecretMaterial(bytes);
CryptographicOperations.ZeroMemory(bytes);
return true;
@@ -69,6 +89,47 @@ internal sealed class EnvironmentSecretProvider : ISecretProvider
return false;
}
}
private static bool TryGetFileSecret(string path, out SecretMaterial? secret)
{
secret = null;
byte[]? bytes = null;
try
{
FileInfo file = new(path);
if (!file.Exists
|| !Path.IsPathFullyQualified(path)
|| file.LinkTarget is not null
|| file.Length is <= 0 or > MaximumSecretBytes)
{
return false;
}
bytes = File.ReadAllBytes(path);
if (bytes.Length is 0 or > MaximumSecretBytes)
{
return false;
}
secret = new SecretMaterial(bytes);
return true;
}
catch (Exception exception) when (exception is IOException
or UnauthorizedAccessException
or ArgumentException
or NotSupportedException
or System.Security.SecurityException)
{
return false;
}
finally
{
if (bytes is not null)
{
CryptographicOperations.ZeroMemory(bytes);
}
}
}
}
internal sealed class EphemeralDevelopmentSecretProvider : ISecretProvider, IDisposable
@@ -142,8 +142,36 @@ internal sealed class SigningKeyRing : IDisposable
return VerificationKeyLookup.Available;
}
public bool Revoke(string keyId) =>
_keys.ContainsKey(keyId) && _runtimeRevocations.TryAdd(keyId, 0);
public bool Revoke(string keyId)
{
if (!_keys.ContainsKey(keyId))
{
return false;
}
_runtimeRevocations.TryAdd(keyId, 0);
return true;
}
public IReadOnlyList<SigningKeyStatus> GetStatuses(DateTimeOffset now) => _keys.Values
.OrderBy(static key => key.KeyId, StringComparer.Ordinal)
.Select(key => new SigningKeyStatus(
key.KeyId,
IsRevoked(key)
? "revoked"
: now < key.NotBefore
? "not-yet-valid"
: now < key.SignUntil
? "signing"
: now < key.VerifyUntil
? "verify-only"
: "retired",
key.SignUntil,
key.VerifyUntil,
key.GameId,
key.EnvironmentId,
key.CredentialKinds.Select(static kind => kind.ToString()).Order().ToArray()))
.ToArray();
public void Dispose()
{
@@ -210,6 +238,15 @@ internal sealed class SigningKeyRing : IDisposable
}
}
internal sealed record SigningKeyStatus(
string KeyId,
string Status,
DateTimeOffset SignUntil,
DateTimeOffset VerifyUntil,
string? GameId,
string? EnvironmentId,
IReadOnlyList<string> CredentialKinds);
internal sealed class SigningKey : IDisposable
{
private byte[]? _material;
@@ -297,6 +297,7 @@ internal sealed record StoredJoinAttempt
public required SecretFingerprint ConnectionTicketFingerprint { get; init; }
public NetworkEndpoint? DedicatedFallback { get; init; }
public required DateTimeOffset ExpiresAt { get; init; }
public required TimeSpan CreatedAtMonotonic { get; init; }
public required DateTimeOffset ConnectionTicketExpiresAt { get; init; }
public AttemptEndpointBinding? HostEndpoint { get; init; }
public AttemptEndpointBinding? ClientEndpoint { get; init; }
@@ -11,19 +11,36 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
private readonly DateTimeOffset _wallOrigin;
private readonly TimeSpan _monotonicOrigin;
private readonly Dictionary<SessionListingId, ListingEntry> _listings = [];
private readonly Dictionary<string, int> _listingCountsByOwner = new(StringComparer.Ordinal);
private readonly PriorityQueue<DeadlineEntry<SessionListingId>, long> _listingExpiries = new();
private readonly HashSet<SessionListingId> _scheduledListingExpiries = [];
private readonly Dictionary<LeaseId, SessionListingId> _leases = [];
private readonly Dictionary<MediationHandle, SessionListingId> _presenceHandles = [];
private readonly Dictionary<MediationHandle, PresenceEntry> _presence = [];
private readonly PriorityQueue<DeadlineEntry<MediationHandle>, long> _presenceExpiries = new();
private readonly HashSet<MediationHandle> _scheduledPresenceExpiries = [];
private readonly Dictionary<JoinAttemptId, AttemptEntry> _attempts = [];
private readonly Dictionary<TenantScope, int> _attemptCountsByScope = [];
private readonly Dictionary<SessionListingId, HashSet<JoinAttemptId>> _attemptsByListing = [];
private readonly PriorityQueue<DeadlineEntry<JoinAttemptId>, long> _attemptExpiries = new();
private readonly Dictionary<JoinAttemptId, OutcomeReportEntry> _outcomeReports = [];
private readonly Dictionary<SessionListingId, HashSet<JoinAttemptId>> _outcomesByListing = [];
private readonly PriorityQueue<DeadlineEntry<JoinAttemptId>, long> _outcomeExpiries = new();
private readonly Dictionary<MediationHandle, JoinAttemptId> _attemptHandles = [];
private readonly Dictionary<string, IdempotencyEntry> _idempotency = new(StringComparer.Ordinal);
private readonly PriorityQueue<DeadlineEntry<string>, long> _idempotencyExpiries = new();
private readonly Dictionary<string, TimeSpan> _replay = new(StringComparer.Ordinal);
private readonly PriorityQueue<DeadlineEntry<string>, long> _replayExpiries = new();
private readonly Dictionary<string, TimeSpan> _revocations = new(StringComparer.Ordinal);
private readonly PriorityQueue<DeadlineEntry<string>, long> _revocationExpiries = new();
private readonly HashSet<string> _scheduledRevocationExpiries = new(StringComparer.Ordinal);
private TimeSpan? _drainDeadline;
private TimeSpan _nextUdpMaintenance;
private long _maintenanceSweepCount;
private long _expiryChurn;
private bool _available = true;
private EphemeralStoreSnapshot? _metricsSnapshot;
private TimeSpan _metricsSnapshotAt = TimeSpan.MinValue;
public InMemoryEphemeralRendezvousStore(
EphemeralStoreOptions options,
@@ -43,6 +60,22 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
public Guid InstanceId { get; }
internal long MaintenanceSweepCount => Interlocked.Read(ref _maintenanceSweepCount);
internal int ScheduledExpiryEntryCount
{
get
{
lock (_gate)
{
return _listingExpiries.Count
+ _presenceExpiries.Count
+ _attemptExpiries.Count
+ _outcomeExpiries.Count
+ _idempotencyExpiries.Count
+ _replayExpiries.Count
+ _revocationExpiries.Count;
}
}
}
public bool IsAvailable
{
@@ -66,6 +99,59 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
}
}
internal EphemeralStoreSnapshot GetSnapshot()
{
lock (_gate)
{
TimeSpan now = _monotonicClock.Elapsed;
Cleanup(now);
EphemeralStoreSnapshot snapshot = CreateSnapshot();
_metricsSnapshot = snapshot;
_metricsSnapshotAt = now;
return snapshot;
}
}
internal int GetActiveJoinAttemptCountForDrain()
{
lock (_gate)
{
_expiryChurn += RemoveExpiredAttempts(_monotonicClock.Elapsed);
return _attempts.Count;
}
}
internal EphemeralStoreSnapshot GetMetricsSnapshot()
{
lock (_gate)
{
TimeSpan now = _monotonicClock.Elapsed;
if (_metricsSnapshot is null
|| now < _metricsSnapshotAt
|| now - _metricsSnapshotAt >= TimeSpan.FromMilliseconds(100))
{
Cleanup(now);
_metricsSnapshot = CreateSnapshot();
_metricsSnapshotAt = now;
}
return _metricsSnapshot;
}
}
private EphemeralStoreSnapshot CreateSnapshot() => new(
_listings.Count,
_presence.Count,
_attempts.Count,
_outcomeReports.Count,
_replay.Count,
_revocations.Count,
_idempotency.Count,
_maintenanceSweepCount,
_expiryChurn,
_available,
_drainDeadline.HasValue);
public StoreResult<StoredListing> CreateListing(
CreateListingCommand command,
CancellationToken cancellationToken = default) => Atomic<StoredListing>(now =>
@@ -104,10 +190,8 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
if (_listings.Count >= _options.MaxListings
|| _idempotency.Count >= _options.MaxIdempotencyEntries
|| _listings.Values.Count(entry => string.Equals(
entry.Definition.OwnerSubject,
command.Listing.OwnerSubject,
StringComparison.Ordinal)) >= command.OwnerListingLimit)
|| _listingCountsByOwner.GetValueOrDefault(command.Listing.OwnerSubject)
>= command.OwnerListingLimit)
{
return new(StoreResultCode.CapacityExceeded);
}
@@ -126,12 +210,21 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
WallDeadline(now, _options.LeaseLifetime),
version: 1);
_listings.Add(frozen.ListingId, entry);
ScheduleMutableDeadline(
_listingExpiries,
_scheduledListingExpiries,
frozen.ListingId,
entry.LeaseDeadline);
_listingCountsByOwner[frozen.OwnerSubject] =
_listingCountsByOwner.GetValueOrDefault(frozen.OwnerSubject) + 1;
_leases.Add(frozen.LeaseId, frozen.ListingId);
_presenceHandles.Add(frozen.HostPresenceHandle, frozen.ListingId);
_idempotency.Add(idempotencyKey, new(
IdempotencyEntry idempotency = new(
command.RequestFingerprint,
frozen.ListingId,
now + _options.IdempotencyLifetime));
now + _options.IdempotencyLifetime);
_idempotency.Add(idempotencyKey, idempotency);
EnqueueDeadline(_idempotencyExpiries, idempotencyKey, idempotency.Deadline);
return new(StoreResultCode.Success, Snapshot(entry));
}, cancellationToken);
@@ -289,10 +382,20 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
return new(StoreResultCode.CapacityExceeded);
}
_presence[command.Handle] = new(
bool isNewPresence = !_presence.ContainsKey(command.Handle);
PresenceEntry presence = new(
command.PublicEndpoint,
command.LocalEndpoint,
now + _options.PresenceLifetime);
_presence[command.Handle] = presence;
if (isNewPresence)
{
ScheduleMutableDeadline(
_presenceExpiries,
_scheduledPresenceExpiries,
command.Handle,
presence.Deadline);
}
return new(StoreResultCode.Success, Snapshot(entry));
}, cancellationToken, eagerCleanup: false);
@@ -387,8 +490,7 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
if (_attempts.Count >= _options.MaxJoinAttempts
|| _outcomeReports.Count >= _options.MaxOutcomeReports
|| _idempotency.Count >= _options.MaxIdempotencyEntries
|| _attempts.Values.Count(entry => entry.Command.Scope == command.Scope)
>= command.ScopeAttemptLimit)
|| _attemptCountsByScope.GetValueOrDefault(command.Scope) >= command.ScopeAttemptLimit)
{
return new(StoreResultCode.CapacityExceeded);
}
@@ -400,19 +502,29 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
AttemptEntry attempt = new(
command,
now,
now + _options.JoinAttemptLifetime,
WallDeadline(now, _options.JoinAttemptLifetime));
_attempts.Add(command.AttemptId, attempt);
_outcomeReports.Add(command.AttemptId, new(
AddToIndex(_attemptsByListing, command.ListingId, command.AttemptId);
_attemptCountsByScope[command.Scope] =
_attemptCountsByScope.GetValueOrDefault(command.Scope) + 1;
EnqueueDeadline(_attemptExpiries, command.AttemptId, attempt.Deadline);
OutcomeReportEntry outcome = new(
command.ListingId,
command.ClientSubject,
command.ClientCapabilityFingerprint,
now + _options.JoinAttemptLifetime + _options.IdempotencyLifetime));
now + _options.JoinAttemptLifetime + _options.IdempotencyLifetime);
_outcomeReports.Add(command.AttemptId, outcome);
AddToIndex(_outcomesByListing, command.ListingId, command.AttemptId);
EnqueueDeadline(_outcomeExpiries, command.AttemptId, outcome.Deadline);
_attemptHandles.Add(command.MediationHandle, command.AttemptId);
_idempotency.Add(idempotencyKey, new(
IdempotencyEntry idempotency = new(
command.RequestFingerprint,
command.AttemptId,
now + _options.IdempotencyLifetime));
now + _options.IdempotencyLifetime);
_idempotency.Add(idempotencyKey, idempotency);
EnqueueDeadline(_idempotencyExpiries, idempotencyKey, idempotency.Deadline);
return new(StoreResultCode.Success, Snapshot(attempt));
}, cancellationToken);
@@ -696,7 +808,9 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
throw new ArgumentOutOfRangeException(nameof(consumption), "Replay lifetime exceeds the configured ceiling.");
}
_replay.Add(key, now + lifetime);
TimeSpan deadline = now + lifetime;
_replay.Add(key, deadline);
EnqueueDeadline(_replayExpiries, key, deadline);
return new(StoreResultCode.Success, true);
}, cancellationToken);
@@ -729,7 +843,28 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
return new(StoreResultCode.CapacityExceeded);
}
_revocations[subject] = now + lifetime;
int activeResourcesBefore = _listings.Count
+ _presence.Count
+ _attempts.Count
+ _outcomeReports.Count;
TimeSpan deadline = now + lifetime;
bool isNewRevocation = !_revocations.TryGetValue(subject, out TimeSpan existingDeadline);
if (!isNewRevocation && existingDeadline > deadline)
{
// A repeated operator action may extend protection but cannot silently
// shorten an already-authoritative security revocation.
deadline = existingDeadline;
}
_revocations[subject] = deadline;
if (isNewRevocation)
{
ScheduleMutableDeadline(
_revocationExpiries,
_scheduledRevocationExpiries,
subject,
deadline);
}
SessionListingId[] listings = _listings
.Where(item => string.Equals(item.Value.Definition.OwnerSubject, subject, StringComparison.Ordinal))
.Select(static item => item.Key)
@@ -753,10 +888,14 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
}
foreach (JoinAttemptId attemptId in outcomeReports)
{
_outcomeReports.Remove(attemptId);
RemoveOutcome(attemptId);
}
return new(StoreResultCode.Success, listings.Length + attempts.Length);
int activeResourcesAfter = _listings.Count
+ _presence.Count
+ _attempts.Count
+ _outcomeReports.Count;
return new(StoreResultCode.Success, activeResourcesBefore - activeResourcesAfter);
}, cancellationToken);
public void BeginDrain(CancellationToken cancellationToken = default)
@@ -768,6 +907,7 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
if (!_drainDeadline.HasValue)
{
_drainDeadline = _monotonicClock.Elapsed + _options.GracefulDrainLifetime;
_metricsSnapshot = null;
}
}
}
@@ -778,6 +918,7 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
{
_available = false;
ClearActiveState();
_metricsSnapshot = null;
}
}
@@ -801,7 +942,9 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
_nextUdpMaintenance = now + UdpMaintenanceInterval;
}
return operation(now);
StoreResult<T> result = operation(now);
_metricsSnapshot = null;
return result;
}
}
@@ -838,60 +981,42 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
ClearActiveState();
}
RemoveExpired(_revocations, now);
RemoveExpired(_replay, now);
foreach (string key in _idempotency
.Where(item => item.Value.Deadline <= now)
.Select(static item => item.Key)
.ToArray())
{
_idempotency.Remove(key);
}
foreach (MediationHandle handle in _presence
.Where(item => item.Value.Deadline <= now)
.Select(static item => item.Key)
.ToArray())
{
_presence.Remove(handle);
}
foreach (JoinAttemptId attemptId in _attempts
.Where(item => item.Value.Deadline <= now)
.Select(static item => item.Key)
.ToArray())
{
RemoveAttempt(attemptId);
}
foreach (JoinAttemptId attemptId in _outcomeReports
.Where(item => item.Value.Deadline <= now)
.Select(static item => item.Key)
.ToArray())
{
_outcomeReports.Remove(attemptId);
}
foreach (SessionListingId listingId in _listings
.Where(item => item.Value.LeaseDeadline <= now)
.Select(static item => item.Key)
.ToArray())
{
RemoveListing(listingId);
}
_expiryChurn += RemoveExpiredMutableDeadlines(
_revocations,
_revocationExpiries,
_scheduledRevocationExpiries,
now);
_expiryChurn += RemoveExpiredDeadlines(_replay, _replayExpiries, now);
_expiryChurn += RemoveExpiredIdempotency(now);
_expiryChurn += RemoveExpiredPresence(now);
_expiryChurn += RemoveExpiredAttempts(now);
_expiryChurn += RemoveExpiredOutcomes(now);
_expiryChurn += RemoveExpiredListings(now);
}
private void ClearActiveState()
{
_listings.Clear();
_listingCountsByOwner.Clear();
_listingExpiries.Clear();
_scheduledListingExpiries.Clear();
_leases.Clear();
_presenceHandles.Clear();
_presence.Clear();
_presenceExpiries.Clear();
_scheduledPresenceExpiries.Clear();
_attempts.Clear();
_attemptCountsByScope.Clear();
_attemptsByListing.Clear();
_attemptExpiries.Clear();
_outcomeReports.Clear();
_outcomesByListing.Clear();
_outcomeExpiries.Clear();
_attemptHandles.Clear();
_idempotency.Clear();
_idempotencyExpiries.Clear();
_replay.Clear();
_replayExpiries.Clear();
}
private void RemoveListing(SessionListingId listingId)
@@ -902,23 +1027,23 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
}
_leases.Remove(listing.Definition.LeaseId);
DecrementCount(_listingCountsByOwner, listing.Definition.OwnerSubject);
_presenceHandles.Remove(listing.Definition.HostPresenceHandle);
_presence.Remove(listing.Definition.HostPresenceHandle);
foreach (JoinAttemptId attemptId in _attempts
.Where(item => item.Value.Command.ListingId == listingId)
.Select(static item => item.Key)
.ToArray())
if (_attemptsByListing.TryGetValue(listingId, out HashSet<JoinAttemptId>? attempts))
{
RemoveAttempt(attemptId);
foreach (JoinAttemptId attemptId in attempts.ToArray())
{
RemoveAttempt(attemptId);
}
}
foreach (JoinAttemptId attemptId in _outcomeReports
.Where(item => item.Value.ListingId == listingId)
.Select(static item => item.Key)
.ToArray())
if (_outcomesByListing.TryGetValue(listingId, out HashSet<JoinAttemptId>? outcomes))
{
_outcomeReports.Remove(attemptId);
foreach (JoinAttemptId attemptId in outcomes.ToArray())
{
RemoveOutcome(attemptId);
}
}
}
@@ -927,9 +1052,260 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
if (_attempts.Remove(attemptId, out AttemptEntry? attempt))
{
_attemptHandles.Remove(attempt.Command.MediationHandle);
DecrementCount(_attemptCountsByScope, attempt.Command.Scope);
RemoveFromIndex(_attemptsByListing, attempt.Command.ListingId, attemptId);
}
}
private void RemoveOutcome(JoinAttemptId attemptId)
{
if (_outcomeReports.Remove(attemptId, out OutcomeReportEntry? outcome))
{
RemoveFromIndex(_outcomesByListing, outcome.ListingId, attemptId);
}
}
private static void AddToIndex<TKey>(
Dictionary<TKey, HashSet<JoinAttemptId>> index,
TKey key,
JoinAttemptId attemptId)
where TKey : notnull
{
if (!index.TryGetValue(key, out HashSet<JoinAttemptId>? values))
{
values = [];
index.Add(key, values);
}
values.Add(attemptId);
}
private static void RemoveFromIndex<TKey>(
Dictionary<TKey, HashSet<JoinAttemptId>> index,
TKey key,
JoinAttemptId attemptId)
where TKey : notnull
{
if (index.TryGetValue(key, out HashSet<JoinAttemptId>? values)
&& values.Remove(attemptId)
&& values.Count == 0)
{
index.Remove(key);
}
}
private static void DecrementCount<TKey>(Dictionary<TKey, int> counts, TKey key)
where TKey : notnull
{
int remaining = counts[key] - 1;
if (remaining == 0)
{
counts.Remove(key);
}
else
{
counts[key] = remaining;
}
}
private int RemoveExpiredAttempts(TimeSpan now)
{
int removed = 0;
while (_attemptExpiries.TryPeek(
out DeadlineEntry<JoinAttemptId> candidate,
out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
_attemptExpiries.Dequeue();
if (_attempts.TryGetValue(candidate.Key, out AttemptEntry? current)
&& current.Deadline == candidate.Deadline)
{
RemoveAttempt(candidate.Key);
removed++;
}
}
return removed;
}
private int RemoveExpiredListings(TimeSpan now)
{
int removed = 0;
while (_listingExpiries.TryPeek(
out DeadlineEntry<SessionListingId> candidate,
out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
_listingExpiries.Dequeue();
_scheduledListingExpiries.Remove(candidate.Key);
if (!_listings.TryGetValue(candidate.Key, out ListingEntry? current))
{
continue;
}
if (current.LeaseDeadline > now)
{
ScheduleMutableDeadline(
_listingExpiries,
_scheduledListingExpiries,
candidate.Key,
current.LeaseDeadline);
}
else
{
RemoveListing(candidate.Key);
removed++;
}
}
return removed;
}
private int RemoveExpiredPresence(TimeSpan now)
{
int removed = 0;
while (_presenceExpiries.TryPeek(
out DeadlineEntry<MediationHandle> candidate,
out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
_presenceExpiries.Dequeue();
_scheduledPresenceExpiries.Remove(candidate.Key);
if (!_presence.TryGetValue(candidate.Key, out PresenceEntry? current))
{
continue;
}
if (current.Deadline > now)
{
ScheduleMutableDeadline(
_presenceExpiries,
_scheduledPresenceExpiries,
candidate.Key,
current.Deadline);
}
else if (_presence.Remove(candidate.Key))
{
removed++;
}
}
return removed;
}
private int RemoveExpiredOutcomes(TimeSpan now)
{
int removed = 0;
while (_outcomeExpiries.TryPeek(
out DeadlineEntry<JoinAttemptId> candidate,
out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
_outcomeExpiries.Dequeue();
if (_outcomeReports.TryGetValue(candidate.Key, out OutcomeReportEntry? current)
&& current.Deadline == candidate.Deadline)
{
RemoveOutcome(candidate.Key);
removed++;
}
}
return removed;
}
private int RemoveExpiredIdempotency(TimeSpan now)
{
int removed = 0;
while (_idempotencyExpiries.TryPeek(
out DeadlineEntry<string> candidate,
out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
_idempotencyExpiries.Dequeue();
if (_idempotency.TryGetValue(candidate.Key, out IdempotencyEntry? current)
&& current.Deadline == candidate.Deadline
&& _idempotency.Remove(candidate.Key))
{
removed++;
}
}
return removed;
}
private static int RemoveExpiredDeadlines<TKey>(
Dictionary<TKey, TimeSpan> entries,
PriorityQueue<DeadlineEntry<TKey>, long> expiries,
TimeSpan now)
where TKey : notnull
{
int removed = 0;
while (expiries.TryPeek(out DeadlineEntry<TKey> candidate, out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
expiries.Dequeue();
if (entries.TryGetValue(candidate.Key, out TimeSpan current)
&& current == candidate.Deadline
&& entries.Remove(candidate.Key))
{
removed++;
}
}
return removed;
}
private static int RemoveExpiredMutableDeadlines<TKey>(
Dictionary<TKey, TimeSpan> entries,
PriorityQueue<DeadlineEntry<TKey>, long> expiries,
HashSet<TKey> scheduled,
TimeSpan now)
where TKey : notnull
{
int removed = 0;
while (expiries.TryPeek(out DeadlineEntry<TKey> candidate, out long deadlineTicks)
&& deadlineTicks <= now.Ticks)
{
expiries.Dequeue();
scheduled.Remove(candidate.Key);
if (!entries.TryGetValue(candidate.Key, out TimeSpan current))
{
continue;
}
if (current > now)
{
ScheduleMutableDeadline(expiries, scheduled, candidate.Key, current);
}
else if (entries.Remove(candidate.Key))
{
removed++;
}
}
return removed;
}
private static void ScheduleMutableDeadline<TKey>(
PriorityQueue<DeadlineEntry<TKey>, long> expiries,
HashSet<TKey> scheduled,
TKey key,
TimeSpan deadline)
where TKey : notnull
{
if (scheduled.Add(key))
{
EnqueueDeadline(expiries, key, deadline);
}
}
private static void EnqueueDeadline<TKey>(
PriorityQueue<DeadlineEntry<TKey>, long> expiries,
TKey key,
TimeSpan deadline)
where TKey : notnull =>
expiries.Enqueue(new(key, deadline), deadline.Ticks);
private bool HandleExists(MediationHandle handle) =>
_presenceHandles.ContainsKey(handle) || _attemptHandles.ContainsKey(handle);
@@ -959,6 +1335,7 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
ClientCapabilityFingerprint = entry.Command.ClientCapabilityFingerprint,
ConnectionTicketFingerprint = entry.Command.ConnectionTicketFingerprint,
DedicatedFallback = StoredListing.CopyEndpoint(entry.Command.DedicatedFallback),
CreatedAtMonotonic = entry.CreatedAtMonotonic,
ExpiresAt = entry.WallExpiresAt,
ConnectionTicketExpiresAt = entry.TicketWallExpiresAt ?? default,
HostEndpoint = entry.HostEndpoint,
@@ -968,17 +1345,6 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
IsCancelled = entry.IsCancelled,
};
private static void RemoveExpired(Dictionary<string, TimeSpan> entries, TimeSpan now)
{
foreach (string key in entries
.Where(item => item.Value <= now)
.Select(static item => item.Key)
.ToArray())
{
entries.Remove(key);
}
}
private static void ValidateListing(ListingDefinition listing)
{
ArgumentNullException.ThrowIfNull(listing);
@@ -1101,10 +1467,12 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
private sealed class AttemptEntry(
CreateJoinAttemptCommand command,
TimeSpan createdAtMonotonic,
TimeSpan deadline,
DateTimeOffset wallExpiresAt)
{
public CreateJoinAttemptCommand Command { get; } = command;
public TimeSpan CreatedAtMonotonic { get; } = createdAtMonotonic;
public SecretFingerprint HostCapabilityFingerprint { get; } = command.HostCapabilityFingerprint;
public SecretFingerprint ClientCapabilityFingerprint { get; } = command.ClientCapabilityFingerprint;
public SecretFingerprint ConnectionTicketFingerprint { get; } = command.ConnectionTicketFingerprint;
@@ -1119,6 +1487,9 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
public bool IsCancelled { get; set; }
}
private readonly record struct DeadlineEntry<TKey>(TKey Key, TimeSpan Deadline)
where TKey : notnull;
private sealed class OutcomeReportEntry(
SessionListingId listingId,
string clientSubject,
@@ -1137,3 +1508,16 @@ internal sealed class InMemoryEphemeralRendezvousStore : IEphemeralRendezvousSto
object ResourceId,
TimeSpan Deadline);
}
internal sealed record EphemeralStoreSnapshot(
int ActiveListings,
int FreshPresenceBindings,
int ActiveJoinAttempts,
int RetainedOutcomeReports,
int ReplayMarkers,
int PrincipalRevocations,
int IdempotencyEntries,
long MaintenanceSweeps,
long ExpiryChurn,
bool IsAvailable,
bool IsDraining);
@@ -1,7 +1,10 @@
using System.Diagnostics;
using System.Net;
using System.Net.Sockets;
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Abuse;
using FinalFactory.Rendezvous.Server.JoinAttempts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.Sessions;
using FinalFactory.Rendezvous.Server.State;
@@ -36,7 +39,10 @@ internal enum NatMediationResult
internal sealed class NatMediationProcessor(
IEphemeralRendezvousStore store,
ISessionCapabilityService capabilities,
JoinAttemptService joinAttempts)
JoinAttemptService joinAttempts,
AbuseProtectionService? abuseProtection = null,
RendezvousTelemetry? telemetry = null,
IMonotonicClock? monotonicClock = null)
{
public NatMediationResult ProcessDatagram(
ReadOnlySpan<byte> encoded,
@@ -44,6 +50,49 @@ internal sealed class NatMediationProcessor(
INatIntroductionSink introductionSink,
CancellationToken cancellationToken = default)
{
if (!TryAcceptIngress(observedPublicEndpoint, "frozen"))
{
return NatMediationResult.Dropped;
}
return ProcessDatagramAfterIngress(
encoded,
observedPublicEndpoint,
introductionSink,
cancellationToken);
}
internal bool TryAcceptIngress(IPEndPoint observedPublicEndpoint, string operation) =>
abuseProtection is null
|| abuseProtection.TryAcceptUdpIngress(observedPublicEndpoint.Address, operation);
internal NatMediationResult ProcessDatagramAfterIngress(
ReadOnlySpan<byte> encoded,
IPEndPoint observedPublicEndpoint,
INatIntroductionSink introductionSink,
CancellationToken cancellationToken = default)
{
long started = Stopwatch.GetTimestamp();
using Activity? activity = telemetry?.StartActivity("UDP frozen", ActivityKind.Server);
NatMediationResult result = ProcessDatagramCore(
encoded,
observedPublicEndpoint,
introductionSink,
cancellationToken);
telemetry?.RecordUdp(
"frozen",
result.ToString(),
Stopwatch.GetElapsedTime(started).TotalMilliseconds);
return result;
}
private NatMediationResult ProcessDatagramCore(
ReadOnlySpan<byte> encoded,
IPEndPoint observedPublicEndpoint,
INatIntroductionSink introductionSink,
CancellationToken cancellationToken)
{
if (!RendezvousUdpCodec.TryDecode(encoded, out PresenceDatagram? datagram, out _)
|| datagram is null
|| datagram.Capability.Length != ContractLimits.DerivedCredentialCharacters
@@ -63,7 +112,7 @@ internal sealed class NatMediationProcessor(
return NatMediationResult.Dropped;
}
NatMediationResult result = ProcessRequest(
NatMediationResult result = ProcessRequestCore(
claimedLocalEndpoint,
observedPublicEndpoint,
NatPunchRequestTokenCodec.Encode(role, datagram.MediationHandle, datagram.Capability),
@@ -76,7 +125,7 @@ internal sealed class NatMediationProcessor(
return result;
}
return ProcessRequest(
return ProcessRequestCore(
claimedLocalEndpoint,
observedPublicEndpoint,
NatPunchRequestTokenCodec.Encode(
@@ -98,6 +147,49 @@ internal sealed class NatMediationProcessor(
ArgumentNullException.ThrowIfNull(observedPublicEndpoint);
ArgumentNullException.ThrowIfNull(introductionSink);
if (!TryAcceptIngress(observedPublicEndpoint, "litenet-or-invalid"))
{
return NatMediationResult.Dropped;
}
return ProcessRequestAfterIngress(
claimedLocalEndpoint,
observedPublicEndpoint,
token,
introductionSink,
cancellationToken);
}
internal NatMediationResult ProcessRequestAfterIngress(
IPEndPoint claimedLocalEndpoint,
IPEndPoint observedPublicEndpoint,
string token,
INatIntroductionSink introductionSink,
CancellationToken cancellationToken = default)
{
long started = Stopwatch.GetTimestamp();
using Activity? activity = telemetry?.StartActivity("UDP litenet", ActivityKind.Server);
NatMediationResult result = ProcessRequestCore(
claimedLocalEndpoint,
observedPublicEndpoint,
token,
introductionSink,
cancellationToken);
telemetry?.RecordUdp(
"litenet",
result.ToString(),
Stopwatch.GetElapsedTime(started).TotalMilliseconds);
return result;
}
private NatMediationResult ProcessRequestCore(
IPEndPoint claimedLocalEndpoint,
IPEndPoint observedPublicEndpoint,
string token,
INatIntroductionSink introductionSink,
CancellationToken cancellationToken)
{
if (!NatPunchRequestTokenCodec.TryDecode(token, out NatPunchRequestToken? request)
|| request is null
|| !TryCreateObservedEndpoint(observedPublicEndpoint, out ObservedEndpoint publicEndpoint)
@@ -106,6 +198,17 @@ internal sealed class NatMediationProcessor(
return NatMediationResult.Dropped;
}
string operation = request.Role.ToString();
if (abuseProtection is not null
&& !abuseProtection.TryAcceptUdpIdentity(
operation,
observedPublicEndpoint.Address,
request.Capability,
request.MediationHandle.ToString()))
{
return NatMediationResult.Dropped;
}
ObservedEndpoint? localEndpoint = TryCreatePrivateCandidate(
claimedLocalEndpoint,
publicEndpoint.AddressFamily,
@@ -186,6 +289,14 @@ internal sealed class NatMediationProcessor(
try
{
introductionSink.Introduce(CreatePlan(consumed.Value, ticket.Value));
if (telemetry is not null && monotonicClock is not null)
{
telemetry.RecordPairingLatency(Math.Max(
0,
(monotonicClock.Elapsed - consumed.Value.Attempt.CreatedAtMonotonic)
.TotalMilliseconds));
}
return NatMediationResult.Introduced;
}
catch (Exception exception) when (exception is SocketException
@@ -8,7 +8,7 @@ using Microsoft.Extensions.Options;
namespace FinalFactory.Rendezvous.Server.Transport;
internal sealed partial class UdpMediatorService : BackgroundService
internal sealed class UdpMediatorService : BackgroundService
{
private readonly ILogger<UdpMediatorService> _logger;
private readonly UdpMediatorOptions _options;
@@ -133,20 +133,23 @@ internal sealed partial class UdpMediatorService : BackgroundService
manager?.Stop();
}
[LoggerMessage(
EventId = 1,
Level = LogLevel.Information,
Message = "UDP mediator listening on {ListenAddress}:{ListenPort}")]
private static partial void LogMediatorListening(
private static readonly Action<ILogger, IPAddress, int, Exception?> MediatorListening =
LoggerMessage.Define<IPAddress, int>(
LogLevel.Information,
new EventId(1, nameof(LogMediatorListening)),
"UDP mediator listening on {ListenAddress}:{ListenPort}");
private static readonly Action<ILogger, Exception?> MediatorStopped = LoggerMessage.Define(
LogLevel.Information,
new EventId(2, nameof(LogMediatorStopped)),
"UDP mediator stopped");
private static void LogMediatorListening(
ILogger logger,
IPAddress listenAddress,
int listenPort);
int listenPort) => MediatorListening(logger, listenAddress, listenPort, null);
[LoggerMessage(
EventId = 2,
Level = LogLevel.Information,
Message = "UDP mediator stopped")]
private static partial void LogMediatorStopped(ILogger logger);
private static void LogMediatorStopped(ILogger logger) => MediatorStopped(logger, null);
private sealed class LiteNetIntroductionSink(NatPunchModule module) : INatIntroductionSink
{
@@ -172,12 +175,21 @@ internal sealed partial class UdpMediatorService : BackgroundService
bool isFrozenEnvelope = length >= 2
&& data[0] == RendezvousUdpCodec.MagicFirst
&& data[1] == RendezvousUdpCodec.MagicSecond;
if (!processor.TryAcceptIngress(
endPoint,
isFrozenEnvelope ? "frozen" : "litenet-or-invalid"))
{
Drop(ref length);
return;
}
INatIntroductionSink? sink = _sink;
if (isFrozenEnvelope)
{
if (sink is not null)
{
_ = processor.ProcessDatagram(data.AsSpan(0, length), endPoint, sink);
_ = processor.ProcessDatagramAfterIngress(
data.AsSpan(0, length), endPoint, sink);
}
}
else if (sink is not null
@@ -188,7 +200,8 @@ internal sealed partial class UdpMediatorService : BackgroundService
&& claimedLocalEndpoint is not null
&& token is not null)
{
_ = processor.ProcessRequest(claimedLocalEndpoint, endPoint, token, sink);
_ = processor.ProcessRequestAfterIngress(
claimedLocalEndpoint, endPoint, token, sink);
}
// Every inbound packet is consumed here. NatPunchModule is used only for outbound introductions.
@@ -1,5 +1,8 @@
{
"Rendezvous": {
"AbuseProtection": {
"OperatorAllowedAddresses": ["127.0.0.1", "::1"]
},
"Provisioning": {
"Issuer": "final-factory-rendezvous-development",
"Audience": "final-factory-rendezvous",
@@ -14,6 +17,14 @@
"NotBefore": "2025-01-01T00:00:00Z",
"SignUntil": "2035-01-01T00:00:00Z",
"VerifyUntil": "2035-01-02T00:00:00Z"
},
{
"KeyId": "development-operator-1",
"SecretReference": "development:ephemeral/rendezvous-operator-signing",
"CredentialKinds": ["Operator"],
"NotBefore": "2025-01-01T00:00:00Z",
"SignUntil": "2035-01-01T00:00:00Z",
"VerifyUntil": "2035-01-02T00:00:00Z"
}
],
"Games": [
@@ -1,10 +1,60 @@
{
"Rendezvous": {
"Deployment": {
"PublicHttpBaseUrl": "",
"PublicUdpHost": "",
"PublicUdpPort": 9050,
"DrainDeadlineSeconds": 30,
"MinimumDrainSeconds": 1,
"SingleActiveInstance": true,
"AllowPrivatePublicEndpoints": false
},
"Udp": {
"ListenAddress": "0.0.0.0",
"Port": 9050,
"MaxDatagramsPerPoll": 256,
"PollIntervalMilliseconds": 2
},
"Audit": {
"MaxEntries": 10000,
"RetentionDays": 30
},
"AbuseProtection": {
"WindowSeconds": 1,
"MaxTrackedKeys": 100000,
"CriticalTrackedKeyReserve": 2048,
"UdpTrackedKeyLimit": 70000,
"TrustedProxyAddresses": [],
"OperatorAllowedAddresses": [],
"HealthGlobalRequestsPerWindow": 1000,
"HealthGlobalConcurrency": 32,
"HealthIpPrefixRequestsPerWindow": 120,
"HealthIpPrefixConcurrency": 8,
"OperatorGlobalRequestsPerWindow": 1000,
"OperatorGlobalConcurrency": 32,
"OperatorIpPrefixRequestsPerWindow": 120,
"OperatorIpPrefixConcurrency": 8,
"HttpGlobalRequestsPerWindow": 20000,
"HttpOptionalRequestsPerWindow": 18000,
"HttpIpPrefixRequestsPerWindow": 500,
"HttpOptionalIpPrefixRequestsPerWindow": 450,
"HttpOperationRequestsPerWindow": 5000,
"HttpTenantRequestsPerWindow": 2000,
"HttpPrincipalRequestsPerWindow": 500,
"HttpResourceRequestsPerWindow": 200,
"HttpGlobalConcurrency": 1024,
"HttpOptionalConcurrency": 768,
"HttpIpPrefixConcurrency": 64,
"HttpOptionalIpPrefixConcurrency": 48,
"HttpOperationConcurrency": 256,
"HttpTenantConcurrency": 256,
"HttpPrincipalConcurrency": 32,
"HttpResourceConcurrency": 16,
"UdpGlobalDatagramsPerWindow": 100000,
"UdpIpPrefixDatagramsPerWindow": 2000,
"UdpOperationDatagramsPerWindow": 50000,
"UdpCapabilityDatagramsPerWindow": 120,
"UdpResourceDatagramsPerWindow": 240
}
},
"Logging": {
@@ -4,7 +4,7 @@
"net10.0": {
"LiteNetLib": {
"type": "Direct",
"requested": "[2.1.4, )",
"requested": "[2.1.4, 2.1.4]",
"resolved": "2.1.4",
"contentHash": "KWlxvMw3Urpqj9joD96LRiK+LC62pQNs/zkXRJc+rHnxgkGp+vV703xzDrxRmv+V1YhCFfIGzs5nrVWtREIlyA=="
},
@@ -0,0 +1,130 @@
using System.Security.Cryptography;
using System.Text;
using LiteNetLib;
using LiteNetLib.Utils;
namespace FinalFactory.Rendezvous.TestClient;
internal sealed class DirectEchoProtocol : IDisposable
{
private const string PingPrefix = "rv1-ping:";
private const string EchoPrefix = "rv1-echo:";
private const string AckPrefix = "rv1-ack:";
private const string DonePrefix = "rv1-done:";
private readonly EventBasedNetListener _events;
private readonly bool _host;
private readonly Dictionary<NetPeer, string> _hostNonces = [];
private readonly TaskCompletionSource<bool> _completed = new(
TaskCreationOptions.RunContinuationsAsynchronously);
private string? _nonce;
private bool _disposed;
internal DirectEchoProtocol(EventBasedNetListener events, bool host)
{
_events = events ?? throw new ArgumentNullException(nameof(events));
_host = host;
_events.NetworkReceiveEvent += OnReceive;
_events.PeerDisconnectedEvent += OnPeerDisconnected;
}
internal Task Completion => _completed.Task;
internal int PendingHostExchangeCount => _hostNonces.Count;
internal event Action<NetPeer>? ExchangeCompleted;
internal void BeginJoin(NetPeer peer)
{
ObjectDisposedException.ThrowIf(_disposed, this);
if (_host || _nonce is not null)
{
throw new InvalidOperationException("The direct echo exchange is already active.");
}
_nonce = Convert.ToHexString(RandomNumberGenerator.GetBytes(16)).ToLowerInvariant();
Send(peer, PingPrefix + _nonce);
}
public void Dispose()
{
if (_disposed)
{
return;
}
_events.NetworkReceiveEvent -= OnReceive;
_events.PeerDisconnectedEvent -= OnPeerDisconnected;
_hostNonces.Clear();
_disposed = true;
}
private void OnReceive(
NetPeer peer,
NetPacketReader reader,
byte channel,
DeliveryMethod deliveryMethod)
{
try
{
ReadOnlySpan<byte> payload = reader.GetRemainingBytesSpan();
if (payload.Length is < 9 or > 64)
{
return;
}
string message = Encoding.ASCII.GetString(payload);
if (_host && TryNonce(message, PingPrefix, out string? pingNonce))
{
_hostNonces[peer] = pingNonce!;
Send(peer, EchoPrefix + pingNonce);
}
else if (_host
&& _hostNonces.TryGetValue(peer, out string? hostNonce)
&& string.Equals(message, AckPrefix + hostNonce, StringComparison.Ordinal))
{
_hostNonces.Remove(peer);
Send(peer, DonePrefix + hostNonce);
ExchangeCompleted?.Invoke(peer);
_completed.TrySetResult(true);
}
else if (!_host
&& _nonce is not null
&& string.Equals(message, EchoPrefix + _nonce, StringComparison.Ordinal))
{
Send(peer, AckPrefix + _nonce);
}
else if (!_host
&& _nonce is not null
&& string.Equals(message, DonePrefix + _nonce, StringComparison.Ordinal))
{
ExchangeCompleted?.Invoke(peer);
_completed.TrySetResult(true);
}
}
finally
{
reader.Recycle();
}
}
private static bool TryNonce(string message, string prefix, out string? nonce)
{
nonce = null;
if (!message.StartsWith(prefix, StringComparison.Ordinal)
|| message.Length != prefix.Length + 32)
{
return false;
}
string candidate = message[prefix.Length..];
if (!candidate.All(static character => character is >= '0' and <= '9'
or >= 'a' and <= 'f'))
{
return false;
}
nonce = candidate;
return true;
}
private static void Send(NetPeer peer, string message) => peer.Send(
Encoding.ASCII.GetBytes(message),
DeliveryMethod.ReliableOrdered);
private void OnPeerDisconnected(NetPeer peer, DisconnectInfo disconnectInfo) =>
_hostNonces.Remove(peer);
}
@@ -0,0 +1,36 @@
using FinalFactory.Rendezvous.Contracts;
namespace FinalFactory.Rendezvous.TestClient;
internal sealed class HostServiceFailureBudget
{
private const int MaximumConsecutiveTransientFailures = 3;
private int _consecutiveTransientFailures;
internal bool ShouldStop(
RendezvousErrorCode error,
DateTimeOffset leaseExpiresAt,
DateTimeOffset now)
{
if (error == RendezvousErrorCode.None)
{
Reset();
return false;
}
if (!IsTransient(error))
{
return true;
}
_consecutiveTransientFailures++;
return _consecutiveTransientFailures >= MaximumConsecutiveTransientFailures
|| now >= leaseExpiresAt;
}
internal void Reset() => _consecutiveTransientFailures = 0;
private static bool IsTransient(RendezvousErrorCode error) => error is
RendezvousErrorCode.RateLimited
or RendezvousErrorCode.ServiceUnavailable
or RendezvousErrorCode.InternalError;
}
@@ -1,16 +1,29 @@
namespace FinalFactory.Rendezvous.TestClient;
/// <summary>
/// Bootstrap entry point for the public-SDK-only diagnostic client.
/// </summary>
public static class Program
{
/// <summary>
/// Runs the bootstrap diagnostic.
/// </summary>
public static int Main()
public static async Task<int> Main(string[] args)
{
Console.WriteLine("Rendezvous TestClient bootstrap is ready.");
return 0;
using CancellationTokenSource shutdown = new();
ConsoleCancelEventHandler cancelHandler = (_, eventArgs) =>
{
eventArgs.Cancel = true;
shutdown.Cancel();
};
Console.CancelKeyPress += cancelHandler;
try
{
TestClientApplication application = new(new RendezvousCommandRunner());
return await application.RunAsync(
args,
Console.In,
Console.Out,
Console.Error,
shutdown.Token).ConfigureAwait(false);
}
finally
{
Console.CancelKeyPress -= cancelHandler;
}
}
}
@@ -0,0 +1,3 @@
using System.Runtime.CompilerServices;
[assembly: InternalsVisibleTo("FinalFactory.Rendezvous.Tests")]
@@ -0,0 +1,59 @@
# FinalFactory.Rendezvous.TestClient
This is a diagnostic executable for exercising Rendezvous through the same public
Client and Contracts API available to a game. It is not a production game client,
server browser, dedicated server, relay, account system, or gameplay host.
The executable has three explicit modes:
- `host` publishes a session, maintains presence and its lease, accepts an
authenticated direct peer, and answers a bounded ping/echo/ack/completion exchange;
- `browse` prints compatible public listings; and
- `join` selects or accepts a listing, drives traversal on its caller-owned
LiteNetLib socket, proves direct traffic, reports the typed outcome, and exits.
Run `dotnet run --project src/FinalFactory.Rendezvous.TestClient -- --help` for
the complete option reference. The repository's
[start-to-finish guide](../../docs/integration/test-client.md) provides an
executable local Compose setup, safe failure drill, JSON automation, and a
phase-by-phase diagnostic table. A typical deployment invocation is:
```bash
export RENDEZVOUS_PUBLISHER_CREDENTIAL='<credential from the deployment boundary>'
dotnet run --project src/FinalFactory.Rendezvous.TestClient -- \
host --service https://rendezvous.example/ --mediator rendezvous.example:9050 \
--game space-game --environment production --region eu-central --protocol 1 \
--script --json --exit-after-echo
```
Publisher credentials are accepted only through a named environment variable.
There is deliberately no command-line credential option because process command
lines are routinely exposed to other local tools and diagnostics. Output uses an
allowlisted event model and never includes lease tokens, punch capabilities,
connection tickets, raw metadata, signing material, or reusable credentials.
Script mode never prompts. Join mode selects the first compatible listing unless
`--listing UUID` fixes the choice. `--json` emits one JSON object per line with
`version: 1`; event names and the process exit codes below are stable automation
contracts. A script-mode host without `--run-seconds` uses `--timeout-seconds` as
its total runtime bound. New optional event properties may be added without changing
the version. JSON help and usage failures are versioned events as well; informational
events use stdout and failures use stderr.
| Exit | Meaning |
|---:|---|
| `0` | Requested diagnostic flow completed successfully |
| `2` | Invalid command or options |
| `3` | Missing or invalid local configuration |
| `10` | HTTP, registration, browser, lease, or socket failure |
| `11` | No compatible session was available or selected |
| `12` | Authorization or traversal reached a typed terminal failure |
| `13` | A requested direct ping/echo proof did not complete |
| `130` | Caller cancellation or Ctrl+C |
The client prints the selected direct endpoint category (`loopback`, `private`, or
`public`) but never the raw endpoint. A traversal failure reports whether an
authoritative dedicated fallback is available; the diagnostic does not connect to
that fallback automatically. A host may publish a policy-authorized endpoint with
`--fallback IP:PORT`. See the repository integration guide for process
orchestration and topology limitations.
@@ -0,0 +1,774 @@
using System.Diagnostics;
using System.Net;
using System.Net.Sockets;
using FinalFactory.Rendezvous.Client;
using FinalFactory.Rendezvous.Contracts;
using LiteNetLib;
namespace FinalFactory.Rendezvous.TestClient;
internal sealed class RendezvousCommandRunner : ITestClientCommandRunner
{
private static readonly TimeSpan PollDelay = TimeSpan.FromMilliseconds(5);
private static readonly TimeSpan HostRefreshInterval = TimeSpan.FromMilliseconds(250);
private static readonly TimeSpan DirectTrafficFlushGrace = TimeSpan.FromMilliseconds(500);
public Task<TestClientExitCode> RunAsync(
TestClientOptions options,
TestClientOutput output,
TextReader input,
CancellationToken cancellationToken) => options.Mode switch
{
TestClientMode.Host => RunHostAsync(options, output, cancellationToken),
TestClientMode.Browse => RunBrowseAsync(options, output, cancellationToken),
TestClientMode.Join => RunJoinAsync(options, output, input, cancellationToken),
_ => Task.FromResult(TestClientExitCode.Usage),
};
private static async Task<TestClientExitCode> RunHostAsync(
TestClientOptions options,
TestClientOutput output,
CancellationToken cancellationToken)
{
string? publisherCredential = Environment.GetEnvironmentVariable(
options.PublisherCredentialEnvironmentVariable);
if (!ContractValidation.IsOpaqueHttpCredentialValid(publisherCredential))
{
output.WriteError(
"host.configuration",
"failed",
"The publisher credential environment variable is missing or invalid.",
phase: "configuration");
return TestClientExitCode.Configuration;
}
string credential = publisherCredential!;
using HttpClient http = CreateHttpClient(options);
RendezvousPublisherClient publisher = new(http, ClientOptions(options));
RendezvousSessionBrowserClient browser = new(http, ClientOptions(options));
RendezvousJoinClient joins = new(http, ClientOptions(options));
RendezvousNetListener events = new();
NetManager manager = events.CreateManager();
if (!manager.Start(options.LocalPort))
{
output.WriteError("host.socket", "failed", "The gameplay UDP socket could not start.", phase: "presence");
return TestClientExitCode.ServiceFailure;
}
PublishedSession? session = null;
using CancellationTokenSource hostOperations = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken);
Task<RendezvousClientResult<int>>? refresh = null;
Task<RendezvousClientResult<RenewLeaseResponse>>? renewal = null;
Task<RendezvousClientResult<GetSessionResponse>>? readiness = null;
DirectEchoProtocol? echo = null;
RendezvousHostCoordinator? coordinator = null;
TestClientExitCode hostResult = TestClientExitCode.ServiceFailure;
bool cleanupFailed = false;
try
{
output.Write("host.registration", "started", phase: "registration");
RendezvousClientResult<PublishedSession> registration;
using (CancellationTokenSource registrationTimeout = CreateOperationTimeout(options, cancellationToken))
{
try
{
registration = await publisher.RegisterAsync(
new RegisterSessionRequest
{
IdempotencyKey = Guid.NewGuid().ToString("N"),
GameId = options.GameId,
EnvironmentId = options.EnvironmentId,
RegionId = options.RegionId,
ProtocolVersion = options.ProtocolVersion,
BuildVersion = options.BuildVersion,
DisplayName = options.DisplayName,
Visibility = ListingVisibility.Public,
Capacity = new SessionCapacity { CurrentPlayers = 1, MaximumPlayers = 8 },
Metadata = new Dictionary<string, string>(options.Metadata, StringComparer.Ordinal),
DedicatedFallback = options.DedicatedFallback,
},
credential,
registrationTimeout.Token).ConfigureAwait(false);
}
catch (OperationCanceledException) when (!cancellationToken.IsCancellationRequested)
{
output.WriteError(
"host.registration",
"timed-out",
"Host registration exceeded the bounded startup stage.",
phase: "registration");
return TestClientExitCode.ServiceFailure;
}
}
if (!registration.IsSuccess || registration.Value is null)
{
WriteServiceFailure(output, "host.registration", "registration", registration);
return TestClientExitCode.ServiceFailure;
}
session = registration.Value;
output.Write(
"host.registered",
"registered",
phase: "registration",
listingId: session.ListingId.ToString(),
displayName: options.DisplayName);
echo = new DirectEchoProtocol(events.GameplayEvents, host: true);
echo.ExchangeCompleted += _ => output.Write(
"host.direct-traffic",
"verified",
phase: "direct-traffic",
endpointType: "peer-to-peer");
coordinator = new RendezvousHostCoordinator(
manager,
events,
options.Mediator,
session,
joins,
CoordinatorOptions(options));
coordinator.AttemptCompleted += (_, completion) =>
{
output.Write(
"host.attempt.completed",
completion.Outcome.IsSuccess ? "connected" : "failed",
phase: completion.Outcome.Phase.ToString(),
outcome: completion.Outcome.Kind.ToString(),
elapsedMilliseconds: ToMilliseconds(completion.Outcome.Elapsed));
if (completion.Outcome.IsSuccess)
{
output.Write(
"host.direct-connect",
"connected",
phase: "direct-connection",
endpointType: "peer-to-peer");
}
};
Stopwatch running = Stopwatch.StartNew();
TimeSpan nextRefresh = TimeSpan.Zero;
TimeSpan nextRenewal = TimeSpan.FromSeconds(session.LeaseRenewAfterSeconds);
TimeSpan nextReadinessProbe = TimeSpan.Zero;
bool directTrafficReported = false;
TimeSpan? directTrafficCompletedAt = null;
bool ready = false;
bool terminalFailure = false;
int previousPendingAttempts = 0;
HostServiceFailureBudget refreshFailures = new();
HostServiceFailureBudget renewalFailures = new();
using PeriodicTimer pollTimer = new(PollDelay);
while (!cancellationToken.IsCancellationRequested)
{
coordinator.Poll();
if (coordinator.State != RendezvousHostState.Active)
{
output.WriteError(
"host.lifecycle",
"failed",
"The host coordinator stopped before shutdown was requested.",
phase: "lifecycle",
outcome: coordinator.State.ToString());
terminalFailure = true;
break;
}
if (coordinator.PendingAttemptCount > previousPendingAttempts)
{
output.Write(
"host.punch",
"started",
phase: "nat-traversal",
count: coordinator.PendingAttemptCount);
}
previousPendingAttempts = coordinator.PendingAttemptCount;
if (readiness is { IsCompleted: true })
{
RendezvousClientResult<GetSessionResponse> result = await readiness.ConfigureAwait(false);
readiness = null;
if (result.IsSuccess)
{
ready = true;
output.Write(
"host.ready",
"ready",
phase: "presence",
listingId: session.ListingId.ToString());
}
else
{
nextReadinessProbe = running.Elapsed + TimeSpan.FromMilliseconds(50);
}
}
if (!ready && readiness is null && running.Elapsed >= nextReadinessProbe)
{
readiness = browser.GetAsync(
session.ListingId,
options.GameId,
options.EnvironmentId,
options.ProtocolVersion,
hostOperations.Token);
}
if (refresh is { IsCompleted: true })
{
RendezvousClientResult<int> result = await refresh.ConfigureAwait(false);
refresh = null;
nextRefresh = running.Elapsed + (result.IsSuccess
? HostRefreshInterval
: TimeSpan.FromSeconds(1));
if (!result.IsSuccess)
{
WriteServiceFailure(output, "host.authorization", "authorization", result);
if (refreshFailures.ShouldStop(result.Error, session.ExpiresAt, DateTimeOffset.UtcNow))
{
terminalFailure = true;
break;
}
}
else
{
refreshFailures.Reset();
}
}
if (refresh is null && running.Elapsed >= nextRefresh)
{
refresh = coordinator.RefreshJoinAttemptsAsync(hostOperations.Token);
}
if (renewal is { IsCompleted: true })
{
RendezvousClientResult<RenewLeaseResponse> result = await renewal.ConfigureAwait(false);
renewal = null;
nextRenewal = running.Elapsed + (result.IsSuccess && result.Value is not null
? TimeSpan.FromSeconds(result.Value.RenewAfterSeconds)
: TimeSpan.FromSeconds(1));
output.Write(
"host.lease",
result.IsSuccess ? "renewed" : "failed",
phase: "lease",
message: result.IsSuccess ? null : SafeServiceMessage(result));
if (!result.IsSuccess
&& renewalFailures.ShouldStop(result.Error, session.ExpiresAt, DateTimeOffset.UtcNow))
{
terminalFailure = true;
break;
}
if (result.IsSuccess)
{
renewalFailures.Reset();
}
}
if (renewal is null && running.Elapsed >= nextRenewal)
{
renewal = publisher.RenewAsync(session, credential, hostOperations.Token);
}
if (echo.Completion.IsCompleted && !directTrafficReported)
{
directTrafficReported = true;
directTrafficCompletedAt = running.Elapsed;
}
if (options.ExitAfterEcho
&& directTrafficCompletedAt.HasValue
&& running.Elapsed - directTrafficCompletedAt.Value >= DirectTrafficFlushGrace)
{
break;
}
if (options.RunDuration.HasValue && running.Elapsed >= options.RunDuration.Value)
{
break;
}
if (!await pollTimer.WaitForNextTickAsync(cancellationToken).ConfigureAwait(false))
{
break;
}
}
if (cancellationToken.IsCancellationRequested)
{
hostResult = TestClientExitCode.Cancelled;
}
else if (terminalFailure)
{
hostResult = TestClientExitCode.ServiceFailure;
}
else if (options.ExitAfterEcho && !directTrafficReported)
{
output.WriteError(
"host.direct-traffic",
"timed-out",
"No authenticated ping/echo/ack exchange completed within the host runtime.",
phase: "direct-traffic");
hostResult = TestClientExitCode.DirectTrafficFailed;
}
else
{
hostResult = TestClientExitCode.Success;
}
}
finally
{
hostOperations.Cancel();
await ObserveCancellationAsync(refresh).ConfigureAwait(false);
await ObserveCancellationAsync(renewal).ConfigureAwait(false);
await ObserveCancellationAsync(readiness).ConfigureAwait(false);
coordinator?.Dispose();
echo?.Dispose();
if (session is not null)
{
using CancellationTokenSource cleanup = new(TimeSpan.FromSeconds(5));
try
{
RendezvousClientResult<bool> deregistered = await publisher.DeregisterAsync(
session,
credential,
cleanup.Token).ConfigureAwait(false);
output.Write(
"host.deregistered",
deregistered.IsSuccess ? "complete" : "failed",
phase: "lifecycle",
listingId: session.ListingId.ToString());
if (!deregistered.IsSuccess)
{
cleanupFailed = true;
}
}
catch (OperationCanceledException)
{
output.WriteError(
"host.deregistered",
"timed-out",
"Deregistration did not complete within the cleanup budget.",
phase: "lifecycle");
cleanupFailed = true;
}
}
manager.Stop();
}
return cleanupFailed && !cancellationToken.IsCancellationRequested
? TestClientExitCode.ServiceFailure
: hostResult;
}
private static async Task<TestClientExitCode> RunBrowseAsync(
TestClientOptions options,
TestClientOutput output,
CancellationToken cancellationToken)
{
using CancellationTokenSource operation = CreateOperationTimeout(options, cancellationToken);
using HttpClient http = CreateHttpClient(options);
RendezvousSessionBrowserClient browser = new(http, ClientOptions(options));
output.Write("browse.sessions", "started", phase: "directory");
RendezvousClientResult<IReadOnlyList<SessionListing>> result = await browser.BrowseAllAsync(
BrowseRequest(options),
maximumPages: 10,
cancellationToken: operation.Token).ConfigureAwait(false);
if (!result.IsSuccess || result.Value is null)
{
WriteServiceFailure(output, "browse.sessions", "directory", result);
return TestClientExitCode.ServiceFailure;
}
WriteListings(output, result.Value);
return result.Value.Count == 0
? TestClientExitCode.NoCompatibleSession
: TestClientExitCode.Success;
}
private static async Task<TestClientExitCode> RunJoinAsync(
TestClientOptions options,
TestClientOutput output,
TextReader input,
CancellationToken cancellationToken)
{
using CancellationTokenSource operation = CreateOperationTimeout(options, cancellationToken);
using HttpClient http = CreateHttpClient(options);
RendezvousSessionBrowserClient browser = new(http, ClientOptions(options));
RendezvousJoinClient joins = new(http, ClientOptions(options));
SessionSelection selection = await SelectListingAsync(
options,
output,
input,
browser,
operation.Token).ConfigureAwait(false);
if (selection.Listing is null)
{
return selection.ExitCode;
}
SessionListing listing = selection.Listing;
RendezvousNetListener events = new();
NetManager manager = events.CreateManager();
if (!manager.Start(options.LocalPort))
{
output.WriteError("join.socket", "failed", "The gameplay UDP socket could not start.", phase: "mediation");
return TestClientExitCode.ServiceFailure;
}
RendezvousClientCoordinator? coordinator = null;
bool directConnected = false;
try
{
output.Write(
"join.authorization",
"started",
phase: "authorization",
listingId: listing.ListingId.ToString());
RendezvousConnectionStartResult start = await joins.CreateConnectionAttemptAsync(
new CreateJoinAttemptRequest
{
IdempotencyKey = Guid.NewGuid().ToString("N"),
GameId = options.GameId,
EnvironmentId = options.EnvironmentId,
ListingId = listing.ListingId,
ProtocolVersion = options.ProtocolVersion,
},
listing.DedicatedFallback,
operation.Token).ConfigureAwait(false);
if (start.Outcome is { } serviceOutcome)
{
cancellationToken.ThrowIfCancellationRequested();
WriteOutcome(output, "join.authorization", serviceOutcome);
WriteFallback(output, serviceOutcome);
return TestClientExitCode.TraversalFailed;
}
CreateJoinAttemptResponse attempt = start.Attempt
?? throw new InvalidOperationException("The typed start result had no attempt or outcome.");
using DirectEchoProtocol echo = new(events.GameplayEvents, host: false);
coordinator = new RendezvousClientCoordinator(
manager,
events,
options.Mediator,
attempt,
CoordinatorOptions(options));
output.Write("join.punch", "started", phase: "nat-traversal");
using (CancellationTokenSource traversal = CreateOperationTimeout(options, cancellationToken))
using (PeriodicTimer traversalPoll = new(PollDelay))
{
RendezvousConnectionState previousState = coordinator.State;
while (!coordinator.IsCompleted)
{
traversal.Token.ThrowIfCancellationRequested();
coordinator.Poll();
if (coordinator.State != previousState)
{
previousState = coordinator.State;
if (previousState == RendezvousConnectionState.Connecting)
{
output.Write(
"join.direct-connect",
"started",
phase: "direct-connection");
}
}
if (!coordinator.IsCompleted
&& !await traversalPoll.WaitForNextTickAsync(traversal.Token).ConfigureAwait(false))
{
break;
}
}
}
RendezvousConnectionOutcome outcome = coordinator.Outcome
?? throw new InvalidOperationException("The completed coordinator had no typed outcome.");
WriteOutcome(output, "join.traversal", outcome);
if (!outcome.IsSuccess || coordinator.ConnectedPeer is null)
{
WriteFallback(output, outcome);
await ReportOutcomeAsync(coordinator, joins, output, cancellationToken).ConfigureAwait(false);
return TestClientExitCode.TraversalFailed;
}
NetPeer peer = coordinator.ConnectedPeer;
directConnected = true;
string endpointType = EndpointType(peer.Address);
output.Write(
"join.connected",
"connected",
phase: "direct-connection",
endpointType: endpointType,
elapsedMilliseconds: ToMilliseconds(outcome.Elapsed));
await ReportOutcomeAsync(coordinator, joins, output, cancellationToken).ConfigureAwait(false);
echo.BeginJoin(peer);
using (CancellationTokenSource traffic = CreateOperationTimeout(options, cancellationToken))
using (PeriodicTimer trafficPoll = new(PollDelay))
{
while (!echo.Completion.IsCompleted)
{
traffic.Token.ThrowIfCancellationRequested();
manager.PollEvents();
if (!echo.Completion.IsCompleted
&& !await trafficPoll.WaitForNextTickAsync(traffic.Token).ConfigureAwait(false))
{
break;
}
}
}
await echo.Completion.ConfigureAwait(false);
output.Write(
"join.direct-traffic",
"verified",
phase: "direct-traffic",
endpointType: endpointType);
peer.Disconnect();
manager.PollEvents();
return TestClientExitCode.Success;
}
catch (OperationCanceledException) when (!cancellationToken.IsCancellationRequested)
{
output.WriteError(
"join.timeout",
"timed-out",
"The bounded join operation timed out.",
phase: "lifecycle",
outcome: ConnectionOutcomeKind.TimedOut.ToString());
if (coordinator is not null && !coordinator.IsCompleted)
{
coordinator.Poll();
}
if (coordinator is not null && !coordinator.IsCompleted)
{
coordinator.Cancel();
coordinator.Poll();
if (coordinator.Outcome is { } timeoutOutcome)
{
WriteOutcome(output, "join.traversal", timeoutOutcome);
WriteFallback(output, timeoutOutcome, listing.DedicatedFallback);
await ReportOutcomeAsync(
coordinator,
joins,
output,
cancellationToken).ConfigureAwait(false);
}
}
return directConnected
? TestClientExitCode.DirectTrafficFailed
: TestClientExitCode.TraversalFailed;
}
finally
{
coordinator?.Dispose();
manager.Stop();
}
}
private static async Task<SessionSelection> SelectListingAsync(
TestClientOptions options,
TestClientOutput output,
TextReader input,
RendezvousSessionBrowserClient browser,
CancellationToken cancellationToken)
{
if (options.ListingId.HasValue)
{
RendezvousClientResult<GetSessionResponse> exact = await browser.GetAsync(
options.ListingId.Value,
options.GameId,
options.EnvironmentId,
options.ProtocolVersion,
cancellationToken).ConfigureAwait(false);
if (!exact.IsSuccess || exact.Value is null)
{
WriteServiceFailure(output, "join.selection", "directory", exact);
return new(null, TestClientExitCode.ServiceFailure);
}
return new(exact.Value.Session, TestClientExitCode.Success);
}
RendezvousClientResult<IReadOnlyList<SessionListing>> result = await browser.BrowseAllAsync(
BrowseRequest(options),
maximumPages: 10,
cancellationToken: cancellationToken).ConfigureAwait(false);
if (!result.IsSuccess || result.Value is null)
{
WriteServiceFailure(output, "join.selection", "directory", result);
return new(null, TestClientExitCode.ServiceFailure);
}
if (result.Value.Count == 0)
{
output.Write("join.selection", "empty", phase: "directory", count: 0);
return new(null, TestClientExitCode.NoCompatibleSession);
}
WriteListings(output, result.Value);
if (options.Script)
{
return new(result.Value[0], TestClientExitCode.Success);
}
output.WritePrompt($"Select session [1-{result.Value.Count}]: ");
string? selection = await input.ReadLineAsync(cancellationToken).ConfigureAwait(false);
SessionListing? selected = int.TryParse(selection, out int index)
&& index >= 1
&& index <= result.Value.Count
? result.Value[index - 1]
: null;
return selected is null
? new(null, TestClientExitCode.NoCompatibleSession)
: new(selected, TestClientExitCode.Success);
}
private static void WriteListings(TestClientOutput output, IReadOnlyList<SessionListing> listings)
{
output.Write("browse.completed", "complete", phase: "directory", count: listings.Count);
foreach (SessionListing listing in listings)
{
output.Write(
"browse.session",
"available",
phase: "directory",
listingId: listing.ListingId.ToString(),
displayName: listing.DisplayName);
}
}
private static BrowseSessionsRequest BrowseRequest(TestClientOptions options) => new()
{
GameId = options.GameId,
EnvironmentId = options.EnvironmentId,
ProtocolVersion = options.ProtocolVersion,
RegionId = options.RegionId,
PageSize = options.PageSize,
ExcludeFull = true,
};
private static HttpClient CreateHttpClient(TestClientOptions options) => new()
{
BaseAddress = options.ServiceUri,
Timeout = Timeout.InfiniteTimeSpan,
};
private static RendezvousClientOptions ClientOptions(TestClientOptions options) => new()
{
RequestTimeout = TimeSpan.FromSeconds(Math.Min(30, options.OperationTimeout.TotalSeconds)),
};
private static RendezvousCoordinatorOptions CoordinatorOptions(TestClientOptions options)
{
TimeSpan phaseTimeout = TimeSpan.FromSeconds(
Math.Min(30, options.OperationTimeout.TotalSeconds * 0.45));
return new RendezvousCoordinatorOptions
{
PunchTimeout = phaseTimeout,
DirectConnectTimeout = phaseTimeout,
};
}
private static CancellationTokenSource CreateOperationTimeout(
TestClientOptions options,
CancellationToken cancellationToken)
{
CancellationTokenSource source = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken);
source.CancelAfter(options.OperationTimeout);
return source;
}
private static async Task ReportOutcomeAsync(
RendezvousClientCoordinator coordinator,
RendezvousJoinClient joins,
TestClientOutput output,
CancellationToken callerCancellationToken)
{
using CancellationTokenSource telemetry = CancellationTokenSource.CreateLinkedTokenSource(
callerCancellationToken);
telemetry.CancelAfter(TimeSpan.FromSeconds(5));
try
{
RendezvousClientResult<ReportConnectionOutcomeResponse> report =
await coordinator.ReportOutcomeAsync(joins, telemetry.Token).ConfigureAwait(false);
output.Write(
"join.outcome-report",
report.IsSuccess ? "accepted" : "failed",
phase: "telemetry",
message: report.IsSuccess ? null : SafeServiceMessage(report));
}
catch (OperationCanceledException) when (!callerCancellationToken.IsCancellationRequested)
{
output.WriteError(
"join.outcome-report",
"cancelled",
"Outcome reporting was cancelled within the operation budget.",
phase: "telemetry");
}
}
private static void WriteOutcome(
TestClientOutput output,
string eventName,
RendezvousConnectionOutcome outcome) => output.Write(
eventName,
outcome.IsSuccess ? "connected" : "failed",
phase: outcome.Phase.ToString(),
outcome: outcome.Kind.ToString(),
elapsedMilliseconds: ToMilliseconds(outcome.Elapsed));
private static void WriteFallback(
TestClientOutput output,
RendezvousConnectionOutcome outcome,
NetworkEndpoint? authoritativeFallback = null)
{
bool hasFallback = outcome.HasDedicatedFallback || authoritativeFallback is not null;
output.Write(
"join.fallback",
hasFallback ? "available" : "unavailable",
phase: "fallback",
outcome: outcome.Kind.ToString(),
endpointType: hasFallback ? "dedicated" : "none");
}
private static void WriteServiceFailure<T>(
TestClientOutput output,
string eventName,
string phase,
RendezvousClientResult<T> result) => output.WriteError(
eventName,
"failed",
SafeServiceMessage(result),
phase,
result.Error.ToString());
private static string SafeServiceMessage<T>(RendezvousClientResult<T> result) =>
$"Rendezvous returned {result.Error}.";
private static async Task ObserveCancellationAsync<T>(Task<T>? task)
{
if (task is null)
{
return;
}
try
{
await task.ConfigureAwait(false);
}
catch (OperationCanceledException)
{
}
catch (ObjectDisposedException)
{
}
}
private static string EndpointType(IPAddress address)
{
if (IPAddress.IsLoopback(address))
{
return "loopback";
}
if (address.AddressFamily == AddressFamily.InterNetworkV6)
{
byte[] ipv6 = address.GetAddressBytes();
return address.IsIPv6LinkLocal || (ipv6[0] & 0xfe) == 0xfc
? "private"
: "public";
}
byte[] bytes = address.GetAddressBytes();
bool privateAddress = bytes[0] == 10
|| bytes[0] == 172 && bytes[1] is >= 16 and <= 31
|| bytes[0] == 192 && bytes[1] == 168;
return privateAddress ? "private" : "public";
}
private static long ToMilliseconds(TimeSpan elapsed) =>
(long)Math.Min(long.MaxValue, Math.Max(0, elapsed.TotalMilliseconds));
private sealed record SessionSelection(
SessionListing? Listing,
TestClientExitCode ExitCode);
}
@@ -0,0 +1,95 @@
namespace FinalFactory.Rendezvous.TestClient;
internal enum TestClientExitCode
{
Success = 0,
Usage = 2,
Configuration = 3,
ServiceFailure = 10,
NoCompatibleSession = 11,
TraversalFailed = 12,
DirectTrafficFailed = 13,
Cancelled = 130,
}
internal interface ITestClientCommandRunner
{
Task<TestClientExitCode> RunAsync(
TestClientOptions options,
TestClientOutput output,
TextReader input,
CancellationToken cancellationToken);
}
internal sealed class TestClientApplication(ITestClientCommandRunner runner)
{
private readonly ITestClientCommandRunner _runner = runner ?? throw new ArgumentNullException(nameof(runner));
internal async Task<int> RunAsync(
string[] args,
TextReader input,
TextWriter standardOutput,
TextWriter standardError,
CancellationToken cancellationToken)
{
bool jsonRequested = args.Contains("--json", StringComparer.Ordinal);
TestClientParseResult parsed = TestClientOptionParser.Parse(args);
TestClientOutput output = new(standardOutput, standardError, jsonRequested);
if (parsed.ShowHelp)
{
if (jsonRequested)
{
output.Write(
"cli.help",
"complete",
phase: "configuration",
message: "Run without --json to read the full command reference.");
}
else
{
await standardOutput.WriteLineAsync(TestClientOptionParser.Usage).ConfigureAwait(false);
}
return (int)TestClientExitCode.Success;
}
if (!parsed.Succeeded || parsed.Options is null)
{
if (jsonRequested)
{
output.WriteError(
"cli.usage",
"failed",
parsed.Error ?? "Invalid command line.",
phase: "configuration");
}
else
{
await standardError.WriteLineAsync(parsed.Error ?? "Invalid command line.").ConfigureAwait(false);
await standardError.WriteLineAsync("Use --help for documented options.").ConfigureAwait(false);
}
return (int)TestClientExitCode.Usage;
}
output = new TestClientOutput(standardOutput, standardError, parsed.Options.Json);
try
{
return (int)await _runner.RunAsync(
parsed.Options,
output,
input,
cancellationToken).ConfigureAwait(false);
}
catch (OperationCanceledException) when (cancellationToken.IsCancellationRequested)
{
output.Write("lifecycle.cancelled", "cancelled", phase: "lifecycle");
return (int)TestClientExitCode.Cancelled;
}
catch (Exception exception)
{
output.WriteError(
"lifecycle.failed",
"failed",
$"Unexpected {exception.GetType().Name}; credentials remain redacted.");
return (int)TestClientExitCode.ServiceFailure;
}
}
}
@@ -0,0 +1,377 @@
using System.Net;
using System.Net.Sockets;
using FinalFactory.Rendezvous.Contracts;
namespace FinalFactory.Rendezvous.TestClient;
internal enum TestClientMode
{
Host,
Browse,
Join,
}
internal sealed class TestClientOptions
{
internal TestClientMode Mode { get; init; }
internal Uri ServiceUri { get; init; } = new("http://127.0.0.1:5000/");
internal IPEndPoint Mediator { get; init; } = new(IPAddress.Loopback, 9050);
internal GameId GameId { get; init; } = new("space-game");
internal EnvironmentId EnvironmentId { get; init; } = new("development");
internal RegionId RegionId { get; init; } = new("local");
internal uint ProtocolVersion { get; init; } = 1;
internal string BuildVersion { get; init; } = "test-client";
internal string DisplayName { get; init; } = "Rendezvous diagnostic host";
internal string PublisherCredentialEnvironmentVariable { get; init; } =
"RENDEZVOUS_PUBLISHER_CREDENTIAL";
internal Dictionary<string, string> Metadata { get; init; } = new(StringComparer.Ordinal);
internal NetworkEndpoint? DedicatedFallback { get; init; }
internal SessionListingId? ListingId { get; init; }
internal int LocalPort { get; init; }
internal int PageSize { get; init; } = 20;
internal TimeSpan OperationTimeout { get; init; } = TimeSpan.FromSeconds(20);
internal TimeSpan? RunDuration { get; init; }
internal bool Script { get; init; }
internal bool Json { get; init; }
internal bool ExitAfterEcho { get; init; }
}
internal sealed class TestClientParseResult
{
private TestClientParseResult(TestClientOptions? options, string? error, bool showHelp)
{
Options = options;
Error = error;
ShowHelp = showHelp;
}
internal TestClientOptions? Options { get; }
internal string? Error { get; }
internal bool ShowHelp { get; }
internal bool Succeeded => Options is not null;
internal static TestClientParseResult Success(TestClientOptions options) => new(options, null, false);
internal static TestClientParseResult Failure(string error) => new(null, error, false);
internal static TestClientParseResult Help() => new(null, null, true);
}
internal static class TestClientOptionParser
{
internal const string Usage = """
Rendezvous diagnostic client
Usage:
rendezvous-test-client host [options]
rendezvous-test-client browse [options]
rendezvous-test-client join [options]
Common options:
--service URL HTTP(S) Rendezvous base URL
--mediator IP:PORT UDP mediator endpoint
--game ID Game scope (default: space-game)
--environment ID Environment scope (default: development)
--protocol NUMBER Exact gameplay protocol (default: 1)
--region ID Region filter/publication (default: local)
--timeout-seconds NUMBER Bounded startup/traversal stage, 1-300 (default: 20)
--port NUMBER Caller-owned gameplay UDP port; 0 chooses one
--page-size NUMBER Bounded browser page size, 1-100 (default: 20)
--script Never prompt; select the first compatible listing
--json Emit one versioned JSON event per line
--help Show this help
Host options:
--publisher-credential-env NAME Environment variable containing the credential
--display-name TEXT Public listing name
--build-version TEXT Public build version
--metadata KEY=VALUE Bounded public metadata; may be repeated
--fallback IP:PORT Optional policy-authorized dedicated fallback
--run-seconds NUMBER Stop after 1-86400 seconds
--exit-after-echo Stop after an authenticated ping/echo/ack exchange
Join options:
--listing UUID Join an exact listing; otherwise browse/select
Credentials are accepted only through the named environment variable. They are never
accepted on the command line and are never written to human or JSON output.
""";
internal static TestClientParseResult Parse(string[] args)
{
if (args.Length == 0 || args.Length == 1 && IsHelp(args[0]))
{
return TestClientParseResult.Help();
}
if (args.Length > 64)
{
return TestClientParseResult.Failure("Too many command-line arguments.");
}
if (!TryMode(args[0], out TestClientMode mode))
{
return TestClientParseResult.Failure("The first argument must be host, browse, or join.");
}
Uri serviceUri = new("http://127.0.0.1:5000/");
IPEndPoint mediator = new(IPAddress.Loopback, 9050);
string game = "space-game";
string environment = "development";
string region = "local";
uint protocol = 1;
string buildVersion = "test-client";
string displayName = "Rendezvous diagnostic host";
string credentialEnvironmentVariable = "RENDEZVOUS_PUBLISHER_CREDENTIAL";
Dictionary<string, string> metadata = new(StringComparer.Ordinal);
NetworkEndpoint? dedicatedFallback = null;
SessionListingId? listingId = null;
int localPort = 0;
int pageSize = 20;
int timeoutSeconds = 20;
int? runSeconds = null;
bool script = false;
bool json = false;
bool exitAfterEcho = false;
HashSet<string> seen = new(StringComparer.Ordinal);
for (int index = 1; index < args.Length; index++)
{
string option = args[index];
if (IsHelp(option))
{
return TestClientParseResult.Help();
}
if (option is "--script" or "--json" or "--exit-after-echo")
{
if (!seen.Add(option))
{
return TestClientParseResult.Failure($"Option {option} was specified more than once.");
}
script |= option == "--script";
json |= option == "--json";
exitAfterEcho |= option == "--exit-after-echo";
continue;
}
if (!option.StartsWith("--", StringComparison.Ordinal)
|| index + 1 >= args.Length)
{
return TestClientParseResult.Failure("Every option must use the form --name value.");
}
string value = args[++index];
if (value.Length is 0 or > 512)
{
return TestClientParseResult.Failure($"Option {option} has an invalid value length.");
}
if (option != "--metadata" && !seen.Add(option))
{
return TestClientParseResult.Failure($"Option {option} was specified more than once.");
}
switch (option)
{
case "--service":
if (!TryServiceUri(value, out serviceUri))
{
return TestClientParseResult.Failure("The service URL must be absolute HTTP(S), credential-free, and query-free.");
}
break;
case "--mediator":
if (!IPEndPoint.TryParse(value, out IPEndPoint? parsedMediator)
|| parsedMediator.Port == 0)
{
return TestClientParseResult.Failure("The mediator must be an IP endpoint with a non-zero port.");
}
mediator = parsedMediator;
break;
case "--game":
game = value;
break;
case "--environment":
environment = value;
break;
case "--region":
region = value;
break;
case "--protocol":
if (!uint.TryParse(value, out protocol) || protocol == 0)
{
return TestClientParseResult.Failure("The protocol must be a positive integer.");
}
break;
case "--build-version":
buildVersion = value;
break;
case "--display-name":
displayName = value;
break;
case "--publisher-credential-env":
if (!IsEnvironmentVariableName(value))
{
return TestClientParseResult.Failure("The credential environment-variable name is invalid.");
}
credentialEnvironmentVariable = value;
break;
case "--metadata":
if (!TryMetadata(value, metadata))
{
return TestClientParseResult.Failure("Metadata must be a unique KEY=VALUE pair with a non-empty key.");
}
break;
case "--fallback":
if (!IPEndPoint.TryParse(value, out IPEndPoint? fallbackEndpoint)
|| fallbackEndpoint.Port == 0)
{
return TestClientParseResult.Failure("The fallback must be an IP endpoint with a non-zero port.");
}
dedicatedFallback = new NetworkEndpoint
{
AddressFamily = fallbackEndpoint.AddressFamily == AddressFamily.InterNetwork
? AddressFamilyKind.Ipv4
: AddressFamilyKind.Ipv6,
Address = fallbackEndpoint.Address.ToString(),
Port = fallbackEndpoint.Port,
};
break;
case "--listing":
if (!Guid.TryParse(value, out Guid parsedListing) || parsedListing == Guid.Empty)
{
return TestClientParseResult.Failure("The listing must be a non-empty UUID.");
}
listingId = new SessionListingId(parsedListing);
break;
case "--port":
if (!int.TryParse(value, out localPort) || localPort is < 0 or > 65_535)
{
return TestClientParseResult.Failure("The local UDP port must be between 0 and 65535.");
}
break;
case "--page-size":
if (!int.TryParse(value, out pageSize)
|| pageSize is < 1 or > ContractLimits.BrowserPageMaxItems)
{
return TestClientParseResult.Failure("The page size is outside the contract limit.");
}
break;
case "--timeout-seconds":
if (!int.TryParse(value, out timeoutSeconds) || timeoutSeconds is < 1 or > 300)
{
return TestClientParseResult.Failure("The timeout must be between 1 and 300 seconds.");
}
break;
case "--run-seconds":
if (!int.TryParse(value, out int parsedRunSeconds)
|| parsedRunSeconds is < 1 or > 86_400)
{
return TestClientParseResult.Failure("The host run duration must be between 1 and 86400 seconds.");
}
runSeconds = parsedRunSeconds;
break;
default:
return TestClientParseResult.Failure($"Unknown option {option}.");
}
}
if (!IsSlug(game, ContractLimits.GameIdMaxCharacters)
|| !IsSlug(environment, ContractLimits.EnvironmentIdMaxCharacters)
|| !IsSlug(region, ContractLimits.RegionIdMaxCharacters)
|| !ContractValidation.IsBuildVersionValid(buildVersion)
|| !ContractValidation.IsDisplayNameValid(displayName)
|| !ContractValidation.IsMetadataValid(metadata))
{
return TestClientParseResult.Failure("One or more game, environment, region, build, or display values violate v1 limits.");
}
if (listingId.HasValue && mode != TestClientMode.Join
|| runSeconds.HasValue && mode != TestClientMode.Host
|| exitAfterEcho && mode != TestClientMode.Host
|| metadata.Count > 0 && mode != TestClientMode.Host
|| dedicatedFallback is not null && mode != TestClientMode.Host
|| seen.Contains("--publisher-credential-env") && mode != TestClientMode.Host
|| seen.Contains("--display-name") && mode != TestClientMode.Host
|| seen.Contains("--build-version") && mode != TestClientMode.Host)
{
return TestClientParseResult.Failure("One or more options do not apply to the selected mode.");
}
return TestClientParseResult.Success(new TestClientOptions
{
Mode = mode,
ServiceUri = serviceUri,
Mediator = mediator,
GameId = new(game),
EnvironmentId = new(environment),
RegionId = new(region),
ProtocolVersion = protocol,
BuildVersion = buildVersion,
DisplayName = displayName,
PublisherCredentialEnvironmentVariable = credentialEnvironmentVariable,
Metadata = metadata,
DedicatedFallback = dedicatedFallback,
ListingId = listingId,
LocalPort = localPort,
PageSize = pageSize,
OperationTimeout = TimeSpan.FromSeconds(timeoutSeconds),
RunDuration = runSeconds.HasValue
? TimeSpan.FromSeconds(runSeconds.Value)
: mode == TestClientMode.Host && script
? TimeSpan.FromSeconds(timeoutSeconds)
: null,
Script = script,
Json = json,
ExitAfterEcho = exitAfterEcho,
});
}
private static bool TryMode(string value, out TestClientMode mode) =>
Enum.TryParse(value, true, out mode) && Enum.IsDefined(mode);
private static bool IsHelp(string value) => value is "--help" or "-h" or "help";
private static bool TryServiceUri(string value, out Uri uri)
{
uri = null!;
if (!Uri.TryCreate(value, UriKind.Absolute, out Uri? parsed)
|| parsed.Scheme is not ("http" or "https")
|| !string.IsNullOrEmpty(parsed.UserInfo)
|| !string.IsNullOrEmpty(parsed.Query)
|| !string.IsNullOrEmpty(parsed.Fragment))
{
return false;
}
UriBuilder builder = new(parsed) { Path = parsed.AbsolutePath.TrimEnd('/') + "/" };
uri = builder.Uri;
return true;
}
private static bool IsEnvironmentVariableName(string value)
{
if (value.Length is 0 or > 64 || !(char.IsLetter(value[0]) || value[0] == '_'))
{
return false;
}
return value.All(static character =>
char.IsAsciiLetterOrDigit(character) || character == '_');
}
private static bool TryMetadata(string value, Dictionary<string, string> metadata)
{
int separator = value.IndexOf('=');
if (separator is < 1 or > ContractLimits.MetadataKeyMaxBytes
|| metadata.Count >= ContractLimits.MetadataMaxKeys)
{
return false;
}
string key = value[..separator];
string metadataValue = value[(separator + 1)..];
return !string.IsNullOrWhiteSpace(key)
&& ContractValidation.IsUtf8LengthWithin(key, ContractLimits.MetadataKeyMaxBytes)
&& ContractValidation.IsUtf8LengthWithin(metadataValue, ContractLimits.MetadataValueMaxBytes)
&& metadata.TryAdd(key, metadataValue);
}
private static bool IsSlug(string value, int maximumCharacters) =>
value.Length is > 0
&& value.Length <= maximumCharacters
&& value[0] is >= 'a' and <= 'z'
&& value.All(static character => character is >= 'a' and <= 'z'
or >= '0' and <= '9'
or '-');
}
@@ -0,0 +1,181 @@
using System.Globalization;
using System.Text;
using System.Text.Json;
namespace FinalFactory.Rendezvous.TestClient;
internal sealed class TestClientOutput(TextWriter standardOutput, TextWriter standardError, bool json)
{
private static readonly JsonSerializerOptions JsonOptions = new(JsonSerializerDefaults.Web)
{
WriteIndented = false,
};
private readonly object _gate = new();
private readonly TextWriter _standardOutput = standardOutput ?? throw new ArgumentNullException(nameof(standardOutput));
private readonly TextWriter _standardError = standardError ?? throw new ArgumentNullException(nameof(standardError));
private readonly bool _json = json;
internal void Write(
string eventName,
string status,
string? phase = null,
string? listingId = null,
string? displayName = null,
string? outcome = null,
string? endpointType = null,
int? count = null,
long? elapsedMilliseconds = null,
string? message = null) => WriteCore(
_standardOutput,
new TestClientEvent
{
Event = SafeToken(eventName) ?? string.Empty,
Status = SafeToken(status) ?? string.Empty,
Phase = SafeToken(phase),
ListingId = SafeToken(listingId),
DisplayName = SafeText(displayName),
Outcome = SafeToken(outcome),
EndpointType = SafeToken(endpointType),
Count = count,
ElapsedMilliseconds = elapsedMilliseconds,
Message = SafeText(message),
});
internal void WriteError(
string eventName,
string status,
string message,
string? phase = null,
string? outcome = null) => WriteCore(
_standardError,
new TestClientEvent
{
Event = SafeToken(eventName) ?? string.Empty,
Status = SafeToken(status) ?? string.Empty,
Phase = SafeToken(phase),
Outcome = SafeToken(outcome),
Message = SafeText(message),
});
internal void WritePrompt(string prompt)
{
if (_json)
{
return;
}
lock (_gate)
{
_standardOutput.Write(SafeText(prompt));
_standardOutput.Flush();
}
}
private void WriteCore(TextWriter writer, TestClientEvent item)
{
string line = _json
? JsonSerializer.Serialize(item, JsonOptions)
: HumanLine(item);
lock (_gate)
{
writer.WriteLine(line);
writer.Flush();
}
}
private static string HumanLine(TestClientEvent item)
{
StringBuilder line = new();
line.Append('[').Append(item.Status).Append("] ").Append(item.Event);
Append(line, "phase", item.Phase);
Append(line, "listing", item.ListingId);
Append(line, "name", item.DisplayName, quote: true);
Append(line, "outcome", item.Outcome);
Append(line, "endpoint", item.EndpointType);
if (item.Count.HasValue)
{
Append(line, "count", item.Count.Value.ToString(System.Globalization.CultureInfo.InvariantCulture));
}
if (item.ElapsedMilliseconds.HasValue)
{
Append(
line,
"elapsedMs",
item.ElapsedMilliseconds.Value.ToString(System.Globalization.CultureInfo.InvariantCulture));
}
Append(line, "message", item.Message, quote: true);
return line.ToString();
}
private static void Append(
StringBuilder builder,
string name,
string? value,
bool quote = false)
{
if (!string.IsNullOrEmpty(value))
{
builder.Append(' ').Append(name).Append('=');
if (quote)
{
builder.Append('"').Append(value.Replace("\\", "\\\\", StringComparison.Ordinal)
.Replace("\"", "\\\"", StringComparison.Ordinal)).Append('"');
}
else
{
builder.Append(value);
}
}
}
private static string? SafeToken(string? value)
{
if (value is null)
{
return null;
}
return new string(value
.Take(96)
.Select(static character => char.IsAsciiLetterOrDigit(character)
|| character is '.' or '-' or '_' or ':'
? char.ToLowerInvariant(character)
: '_')
.ToArray());
}
private static string? SafeText(string? value)
{
if (value is null)
{
return null;
}
return new string(value
.Take(160)
.Select(static character => IsUnsafeHumanCharacter(character) ? '?' : character)
.ToArray());
}
private static bool IsUnsafeHumanCharacter(char character) =>
char.GetUnicodeCategory(character) is
UnicodeCategory.Control
or UnicodeCategory.Format
or UnicodeCategory.LineSeparator
or UnicodeCategory.ParagraphSeparator
or UnicodeCategory.Surrogate
or UnicodeCategory.PrivateUse;
private sealed class TestClientEvent
{
public int Version { get; init; } = 1;
public string Event { get; init; } = string.Empty;
public string Status { get; init; } = string.Empty;
public string? Phase { get; init; }
public string? ListingId { get; init; }
public string? DisplayName { get; init; }
public string? Outcome { get; init; }
public string? EndpointType { get; init; }
public int? Count { get; init; }
public long? ElapsedMilliseconds { get; init; }
public string? Message { get; init; }
}
}
@@ -4,7 +4,7 @@
"net8.0": {
"LiteNetLib": {
"type": "Direct",
"requested": "[2.1.4, )",
"requested": "[2.1.4, 2.1.4]",
"resolved": "2.1.4",
"contentHash": "KWlxvMw3Urpqj9joD96LRiK+LC62pQNs/zkXRJc+rHnxgkGp+vV703xzDrxRmv+V1YhCFfIGzs5nrVWtREIlyA=="
},
@@ -22,7 +22,7 @@
"type": "Project",
"dependencies": {
"FinalFactory.Rendezvous.Contracts": "[1.0.0, )",
"LiteNetLib": "[2.1.4, )"
"LiteNetLib": "[2.1.4, 2.1.4]"
}
},
"finalfactory.rendezvous.contracts": {
@@ -0,0 +1,91 @@
namespace FinalFactory.Rendezvous.Capacity;
internal sealed record CapacityOptions
{
public required string Profile { get; init; }
public required int Listings { get; init; }
public required int Attempts { get; init; }
public required int Samples { get; init; }
public required int SoakCycles { get; init; }
public required int SoakSeconds { get; init; }
public string? OutputPath { get; init; }
public static CapacityOptions Parse(string[] args)
{
Dictionary<string, string> values = ParseArguments(args);
string profile = values.GetValueOrDefault("--profile") ?? "quick";
(int listings, int attempts, int samples, int soakCycles, int soakSeconds) = profile switch
{
"quick" => (1_000, 500, 100, 20, 0),
"candidate" => (25_000, 10_000, 1_000, 1_000, 300),
_ => throw new ArgumentException("--profile must be 'quick' or 'candidate'."),
};
return new()
{
Profile = profile,
Listings = ParsePositive(values.GetValueOrDefault("--listings"), listings, "--listings"),
Attempts = ParsePositive(values.GetValueOrDefault("--attempts"), attempts, "--attempts"),
Samples = ParsePositive(values.GetValueOrDefault("--samples"), samples, "--samples"),
SoakCycles = ParsePositive(
values.GetValueOrDefault("--soak-cycles"),
soakCycles,
"--soak-cycles"),
SoakSeconds = ParseNonNegative(
values.GetValueOrDefault("--soak-seconds"),
soakSeconds,
"--soak-seconds"),
OutputPath = values.GetValueOrDefault("--output"),
};
}
private static Dictionary<string, string> ParseArguments(string[] args)
{
HashSet<string> allowed =
[
"--profile",
"--listings",
"--attempts",
"--samples",
"--soak-cycles",
"--soak-seconds",
"--output",
];
Dictionary<string, string> values = new(StringComparer.Ordinal);
for (int index = 0; index < args.Length; index += 2)
{
string option = args[index];
if (!allowed.Contains(option))
{
throw new ArgumentException($"Unknown option: {option}.");
}
if (index == args.Length - 1
|| args[index + 1].StartsWith("--", StringComparison.Ordinal))
{
throw new ArgumentException($"{option} requires a value.");
}
if (!values.TryAdd(option, args[index + 1]))
{
throw new ArgumentException($"{option} may be supplied only once.");
}
}
return values;
}
private static int ParsePositive(string? value, int fallback, string option) =>
value is null
? fallback
: int.TryParse(value, out int parsed) && parsed > 0
? parsed
: throw new ArgumentException($"{option} must be a positive integer.");
private static int ParseNonNegative(string? value, int fallback, string option) =>
value is null
? fallback
: int.TryParse(value, out int parsed) && parsed >= 0
? parsed
: throw new ArgumentException($"{option} must be a non-negative integer.");
}
@@ -0,0 +1,76 @@
namespace FinalFactory.Rendezvous.Capacity;
internal sealed record CapacityReport
{
public required int SchemaVersion { get; init; }
public required string EvidenceVersion { get; init; }
public required DateTimeOffset GeneratedAt { get; init; }
public required string Profile { get; init; }
public required RuntimeEvidence Runtime { get; init; }
public required CapacityTargets Targets { get; init; }
public required IReadOnlyList<CapacityMeasurement> Measurements { get; init; }
public required StateEvidence State { get; init; }
public required IReadOnlyList<string> Failures { get; init; }
public required bool Passed { get; init; }
}
internal sealed record RuntimeEvidence(
string Framework,
string OperatingSystem,
string Kernel,
string Architecture,
string CpuModel,
int ProcessorCount,
string CpuAffinity,
string CpuQuota,
string MemoryLimit,
string GarbageCollector,
string CommitSha,
string TreeState,
string Command,
string ImageDigest,
string WorkloadSeed,
double CapacityPhaseAverageCpuPercent,
long PeakWorkingSetBytes,
long ManagedBytesAfterCleanup);
internal sealed record CapacityTargets(
int VisibleListings,
int ActiveJoinAttempts,
int CoreControlOperationsPerSecond,
int CoreMediationOperationsPerSecond,
double CoreControlP95Milliseconds,
double CoreMediationP95Milliseconds,
double MaximumAverageCpuPercent,
long MaximumWorkingSetBytes,
int SoakCycles,
int SoakDurationSeconds);
internal sealed record CapacityMeasurement(
string Operation,
int Samples,
double P50Milliseconds,
double P95Milliseconds,
double P99Milliseconds,
double OperationsPerSecond,
double MinimumOperationsPerSecond,
double BudgetMilliseconds,
bool Passed);
internal sealed record StateEvidence(
int PeakListings,
int PeakAttempts,
int PeakReplayMarkers,
int FinalListings,
int FinalAttempts,
int FinalReplayMarkers,
long ExpiryChurn,
long MaintenanceSweeps,
int SoakCyclesCompleted,
double SoakDurationSeconds,
int SoakPeakScheduledExpiryEntries,
long SoakManagedGrowthBytes,
int SoakHandleGrowth,
bool RestartStartedEmpty,
bool OverloadWasTyped,
bool RecoverySucceeded);
@@ -0,0 +1,580 @@
using System.Collections.Concurrent;
using System.Diagnostics;
using System.Runtime;
using System.Runtime.InteropServices;
using FinalFactory.Rendezvous.Contracts;
using FinalFactory.Rendezvous.Server.Observability;
using FinalFactory.Rendezvous.Server.State;
namespace FinalFactory.Rendezvous.Capacity;
internal static class CapacityRunner
{
private static readonly TenantScope Scope = new(new("space-game"), new("production"));
private const uint ProtocolVersion = 1;
public static Task<CapacityReport> RunAsync(CapacityOptions options)
{
ArgumentNullException.ThrowIfNull(options);
Process process = Process.GetCurrentProcess();
TimeSpan cpuBefore = process.TotalProcessorTime;
Stopwatch capacityPhaseTime = Stopwatch.StartNew();
List<string> failures = [];
List<CapacityMeasurement> measurements = [];
ManualClock clock = new();
InMemoryEphemeralRendezvousStore store = CreateStore(options, clock);
List<StoredListing> listings = new(options.Listings);
List<CreateJoinAttemptCommand> attempts = new(options.Attempts);
int registrationSamples = Math.Min(options.Samples, options.Listings);
int attemptSamples = Math.Min(options.Samples, options.Attempts);
for (int index = 0; index < options.Listings - registrationSamples; index++)
{
listings.Add(CreateVisibleListing(store, index));
}
measurements.Add(Measure(
"registration-and-presence",
registrationSamples,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 200,
index => listings.Add(CreateVisibleListing(
store,
options.Listings - registrationSamples + index))));
measurements.Add(Measure(
"lease-renewal",
registrationSamples,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 200,
index =>
{
StoredListing listing = listings[index];
StoreResult<StoredListing> renewed = store.RenewLease(new(
listing.Definition.ListingId,
listing.Definition.LeaseId,
listing.Definition.LeaseFingerprint,
listing.Definition.OwnerSubject,
listing.Version));
RequireSuccess(renewed, "renewal");
listings[index] = renewed.Value!;
}));
int browseSamples = Math.Min(options.Samples, 250);
measurements.Add(Measure(
"visible-session-browse",
browseSamples,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 200,
_ => RequireSuccess(
store.BrowseVisibleListings(new(
Scope,
ProtocolVersion,
new RegionId("eu-central"),
ContractLimits.BrowserPageMaxItems,
ExcludeFull: true)),
"browse")));
for (int index = 0; index < options.Attempts - attemptSamples; index++)
{
CreateJoinAttemptCommand command = CreateAttempt(index, listings[index % listings.Count]);
RequireSuccess(store.CreateJoinAttempt(command), "join issuance");
attempts.Add(command);
}
measurements.Add(Measure(
"join-attempt-issuance",
attemptSamples,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 200,
index =>
{
int sequence = options.Attempts - attemptSamples + index;
CreateJoinAttemptCommand command = CreateAttempt(
sequence,
listings[sequence % listings.Count]);
RequireSuccess(store.CreateJoinAttempt(command), "join issuance");
attempts.Add(command);
}));
int punchSamples = Math.Min(options.Samples, attempts.Count);
measurements.Add(MeasureConcurrentPunch(store, attempts, punchSamples));
EphemeralStoreSnapshot peak = store.GetSnapshot();
StoreResult<StoredJoinAttempt> overloaded = store.CreateJoinAttempt(
CreateAttempt(options.Attempts + 1, listings[^1]));
bool overloadWasTyped = peak.ActiveJoinAttempts == options.Attempts
&& overloaded.Code == StoreResultCode.CapacityExceeded;
if (!overloadWasTyped)
{
failures.Add(
$"Expected typed CapacityExceeded at {options.Attempts} active attempts, "
+ $"observed count={peak.ActiveJoinAttempts}, result={overloaded.Code}.");
}
int revocationSamples = Math.Min(Math.Max(1, options.Samples / 20), listings.Count / 2);
measurements.Add(Measure(
"principal-revocation",
revocationSamples,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 50,
index => RequireSuccess(
store.RevokePrincipal(
listings[index].Definition.OwnerSubject,
TimeSpan.FromMinutes(1)),
"principal revocation")));
using (RendezvousTelemetry telemetry = new(store))
{
measurements.Add(Measure(
"telemetry-recording",
Math.Max(100, options.Samples),
budgetMilliseconds: 1,
minimumOperationsPerSecond: 10_000,
_ =>
{
telemetry.RecordHttp("browse", 200, 1);
telemetry.RecordUdp("contribution", "accepted", 1);
telemetry.RecordPairingLatency(2);
}));
}
clock.Advance(TimeSpan.FromSeconds(61));
EphemeralStoreSnapshot? afterCoincidentExpiry = null;
measurements.Add(Measure(
"coincident-listing-attempt-expiry",
samples: 1,
budgetMilliseconds: 200,
minimumOperationsPerSecond: 0,
_ => afterCoincidentExpiry = store.GetSnapshot()));
if (afterCoincidentExpiry!.ActiveJoinAttempts != 0
|| afterCoincidentExpiry.ActiveListings != 0)
{
failures.Add("Coincident 60-second cleanup retained expired listings or join attempts.");
}
clock.Advance(TimeSpan.FromSeconds(90));
_ = store.GetSnapshot();
StoredListing recoveryListing = CreateVisibleListing(store, options.Listings + 1);
StoreResult<StoredJoinAttempt> recovered = store.CreateJoinAttempt(
CreateAttempt(options.Attempts + 2, recoveryListing));
bool recoverySucceeded = recovered.Succeeded;
if (!recoverySucceeded)
{
failures.Add($"Store did not recover after attempt expiry: {recovered.Code}.");
}
clock.Advance(TimeSpan.FromSeconds(151));
EphemeralStoreSnapshot final = store.GetSnapshot();
if (final.ActiveListings != 0
|| final.ActiveJoinAttempts != 0
|| final.ReplayMarkers != 0
|| final.IdempotencyEntries != 0
|| final.RetainedOutcomeReports != 0)
{
failures.Add("Expiry cleanup left active or retained state after every configured deadline.");
}
capacityPhaseTime.Stop();
process.Refresh();
double capacityPhaseCpuPercent = 100
* (process.TotalProcessorTime - cpuBefore).TotalSeconds
/ Math.Max(capacityPhaseTime.Elapsed.TotalSeconds * Environment.ProcessorCount, 0.000_001);
if (options.Profile == "candidate" && capacityPhaseCpuPercent > 70)
{
failures.Add(
$"Capacity-phase CPU {capacityPhaseCpuPercent:F1}% exceeded the 70% candidate budget.");
}
SoakEvidence soak = RunAcceleratedSoak(options, failures);
ManualClock restartClock = new();
EphemeralStoreSnapshot restarted = CreateStore(options, restartClock).GetSnapshot();
bool restartStartedEmpty = restarted.ActiveListings == 0
&& restarted.ActiveJoinAttempts == 0
&& restarted.ReplayMarkers == 0;
if (!restartStartedEmpty)
{
failures.Add("A restarted store did not begin empty.");
}
foreach (CapacityMeasurement measurement in measurements.Where(static item => !item.Passed))
{
failures.Add(
$"{measurement.Operation} missed its budget: p95={measurement.P95Milliseconds:F3} ms, "
+ $"rate={measurement.OperationsPerSecond:F1}/s.");
}
GC.Collect();
GC.WaitForPendingFinalizers();
GC.Collect();
process.Refresh();
long managedAfterCleanup = GC.GetTotalMemory(forceFullCollection: true);
long memoryBudget = 1_610_612_736;
if (process.PeakWorkingSet64 > memoryBudget)
{
failures.Add(
$"Peak working set {process.PeakWorkingSet64} exceeded the 1.5 GiB profile budget.");
}
if (options.Profile == "candidate" && Environment.ProcessorCount != 2)
{
failures.Add(
$"Candidate evidence must expose exactly two CPUs; runtime exposed "
+ $"{Environment.ProcessorCount}.");
}
CapacityReport report = new()
{
SchemaVersion = 2,
EvidenceVersion = "v2",
GeneratedAt = DateTimeOffset.UtcNow,
Profile = options.Profile,
Runtime = new(
RuntimeInformation.FrameworkDescription,
RuntimeInformation.OSDescription,
Environment.OSVersion.VersionString,
RuntimeInformation.ProcessArchitecture.ToString(),
ReadCpuModel(),
Environment.ProcessorCount,
Environment.GetEnvironmentVariable("RENDEZVOUS_EVIDENCE_CPUSET") ?? "unrestricted",
ReadCgroupValue("/sys/fs/cgroup/cpu.max"),
ReadCgroupValue("/sys/fs/cgroup/memory.max"),
GCSettings.IsServerGC ? "server" : "workstation",
Environment.GetEnvironmentVariable("RENDEZVOUS_EVIDENCE_COMMIT") ?? "unrecorded",
Environment.GetEnvironmentVariable("RENDEZVOUS_EVIDENCE_TREE_STATE") ?? "unrecorded",
Environment.GetEnvironmentVariable("RENDEZVOUS_EVIDENCE_COMMAND") ?? "unrecorded",
Environment.GetEnvironmentVariable("RENDEZVOUS_EVIDENCE_IMAGE_DIGEST") ?? "not-containerized",
"fixed-sequences-random-identifiers",
capacityPhaseCpuPercent,
process.PeakWorkingSet64,
managedAfterCleanup),
Targets = new(
options.Listings,
options.Attempts,
200,
2_000,
200,
100,
70,
memoryBudget,
options.SoakCycles,
options.SoakSeconds),
Measurements = measurements,
State = new(
peak.ActiveListings,
peak.ActiveJoinAttempts,
peak.ReplayMarkers,
final.ActiveListings,
final.ActiveJoinAttempts,
final.ReplayMarkers,
final.ExpiryChurn,
final.MaintenanceSweeps,
soak.Cycles,
soak.Duration.TotalSeconds,
soak.PeakScheduledExpiryEntries,
soak.ManagedGrowthBytes,
soak.HandleGrowth,
restartStartedEmpty,
overloadWasTyped,
recoverySucceeded),
Failures = failures,
Passed = failures.Count == 0,
};
return Task.FromResult(report);
}
private static InMemoryEphemeralRendezvousStore CreateStore(
CapacityOptions options,
ManualClock clock) => new(
new EphemeralStoreOptions
{
MaxListings = options.Listings,
MaxPresenceBindings = options.Listings,
MaxJoinAttempts = options.Attempts,
MaxOutcomeReports = options.Attempts,
MaxIdempotencyEntries = options.Listings + options.Attempts + 1,
},
clock,
clock);
private static StoredListing CreateVisibleListing(
InMemoryEphemeralRendezvousStore store,
int sequence)
{
string owner = $"publisher-{sequence}";
SecretFingerprint leaseFingerprint = new($"lease-{sequence}");
SecretFingerprint presenceFingerprint = new($"presence-{sequence}");
ListingDefinition definition = new()
{
ListingId = new(Guid.NewGuid()),
LeaseId = new(Guid.NewGuid()),
Scope = Scope,
OwnerSubject = owner,
RegionId = new("eu-central"),
ProtocolVersion = ProtocolVersion,
BuildVersion = "1.0.0",
DisplayName = $"Capacity host {sequence}",
Visibility = ListingVisibility.Public,
TrustMode = PublisherTrustMode.ManagedDedicated,
CurrentPlayers = 1,
MaximumPlayers = 8,
Metadata = new Dictionary<string, string>(StringComparer.Ordinal),
LeaseFingerprint = leaseFingerprint,
HostPresenceHandle = new(Guid.NewGuid()),
HostPresenceFingerprint = presenceFingerprint,
CapabilityDerivationSalt = "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA",
};
StoreResult<StoredListing> created = store.CreateListing(new(
$"register-{sequence}",
$"register-request-{sequence}",
definition));
RequireSuccess(created, "registration");
StoreResult<StoredListing> bound = store.BindHostPresence(new(
definition.HostPresenceHandle,
presenceFingerprint,
PublicEndpoint(10_000 + sequence % 50_000),
null));
RequireSuccess(bound, "host presence");
return bound.Value!;
}
private static CreateJoinAttemptCommand CreateAttempt(int sequence, StoredListing listing) => new()
{
IdempotencyOwner = $"client-{sequence}",
IdempotencyKey = $"join-{sequence}",
RequestFingerprint = $"join-request-{sequence}",
ClientSubject = $"client-{sequence}",
AttemptId = new(Guid.NewGuid()),
MediationHandle = new(Guid.NewGuid()),
Scope = Scope,
ListingId = listing.Definition.ListingId,
ProtocolVersion = ProtocolVersion,
HostCapabilityFingerprint = new($"host-capability-{sequence}"),
ClientCapabilityFingerprint = new($"client-capability-{sequence}"),
ConnectionTicketFingerprint = new($"ticket-{sequence}"),
CapabilityDerivationSalt = "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA",
};
private static CapacityMeasurement MeasureConcurrentPunch(
InMemoryEphemeralRendezvousStore store,
List<CreateJoinAttemptCommand> attempts,
int samples)
{
ConcurrentBag<double> latencies = [];
Stopwatch total = Stopwatch.StartNew();
Parallel.ForEach(
Enumerable.Range(0, samples),
new ParallelOptions { MaxDegreeOfParallelism = Math.Min(64, Environment.ProcessorCount * 4) },
index =>
{
CreateJoinAttemptCommand attempt = attempts[index];
Stopwatch elapsed = Stopwatch.StartNew();
RequireSuccess(store.BindAttemptEndpoint(new(
attempt.MediationHandle,
AttemptPeerRole.Host,
attempt.HostCapabilityFingerprint,
PublicEndpoint(20_000 + index % 20_000),
null)), "host punch");
RequireSuccess(store.BindAttemptEndpoint(new(
attempt.MediationHandle,
AttemptPeerRole.Client,
attempt.ClientCapabilityFingerprint,
PublicEndpoint(40_000 + index % 20_000),
null)), "client punch");
RequireSuccess(store.ConsumeIntroduction(attempt.MediationHandle), "introduction");
latencies.Add(elapsed.Elapsed.TotalMilliseconds);
});
total.Stop();
return BuildMeasurement(
"simultaneous-punch-pairing",
latencies.ToArray(),
total.Elapsed,
budgetMilliseconds: 100,
minimumOperationsPerSecond: 2_000);
}
private static CapacityMeasurement Measure(
string operation,
int samples,
double budgetMilliseconds,
double minimumOperationsPerSecond,
Action<int> action)
{
double[] latencies = new double[samples];
Stopwatch total = Stopwatch.StartNew();
for (int index = 0; index < samples; index++)
{
long started = Stopwatch.GetTimestamp();
action(index);
latencies[index] = Stopwatch.GetElapsedTime(started).TotalMilliseconds;
}
total.Stop();
return BuildMeasurement(
operation,
latencies,
total.Elapsed,
budgetMilliseconds,
minimumOperationsPerSecond);
}
private static CapacityMeasurement BuildMeasurement(
string operation,
double[] latencies,
TimeSpan elapsed,
double budgetMilliseconds,
double minimumOperationsPerSecond)
{
Array.Sort(latencies);
double operationsPerSecond = latencies.Length / Math.Max(elapsed.TotalSeconds, 0.000_001);
double p95 = Percentile(latencies, 0.95);
return new(
operation,
latencies.Length,
Percentile(latencies, 0.50),
p95,
Percentile(latencies, 0.99),
operationsPerSecond,
minimumOperationsPerSecond,
budgetMilliseconds,
p95 <= budgetMilliseconds && operationsPerSecond >= minimumOperationsPerSecond);
}
private static double Percentile(double[] sorted, double percentile)
{
int index = Math.Clamp((int)Math.Ceiling(sorted.Length * percentile) - 1, 0, sorted.Length - 1);
return sorted[index];
}
private static SoakEvidence RunAcceleratedSoak(
CapacityOptions options,
List<string> failures)
{
GC.Collect();
GC.WaitForPendingFinalizers();
GC.Collect();
long managedBefore = GC.GetTotalMemory(forceFullCollection: true);
int handlesBefore = Process.GetCurrentProcess().HandleCount;
ManualClock clock = new();
CapacityOptions soakOptions = options with { Listings = 100, Attempts = 100 };
InMemoryEphemeralRendezvousStore store = CreateStore(soakOptions, clock);
Stopwatch elapsed = Stopwatch.StartNew();
int cycle = 0;
int peakScheduledExpiryEntries = 0;
while (cycle < options.SoakCycles
|| elapsed.Elapsed < TimeSpan.FromSeconds(options.SoakSeconds))
{
StoredListing listing = CreateVisibleListing(store, cycle);
for (int refresh = 0; refresh < 10; refresh++)
{
clock.Advance(TimeSpan.FromTicks(1));
listing = RequireSuccess(store.RenewLease(new(
listing.Definition.ListingId,
listing.Definition.LeaseId,
listing.Definition.LeaseFingerprint,
listing.Definition.OwnerSubject,
listing.Version)), "soak lease refresh");
listing = RequireSuccess(store.BindHostPresence(new(
listing.Definition.HostPresenceHandle,
listing.Definition.HostPresenceFingerprint,
PublicEndpoint(10_000 + cycle % 50_000),
null)), "soak presence refresh");
}
CreateJoinAttemptCommand attempt = CreateAttempt(cycle, listing);
RequireSuccess(store.CreateJoinAttempt(attempt), "soak join issuance");
RequireSuccess(store.ConsumeReplay(new("capacity-soak", $"replay-{cycle}")), "soak replay");
peakScheduledExpiryEntries = Math.Max(
peakScheduledExpiryEntries,
store.ScheduledExpiryEntryCount);
if (store.ScheduledExpiryEntryCount > 7)
{
failures.Add(
$"Mutable deadline refresh grew the expiry queue to "
+ $"{store.ScheduledExpiryEntryCount} entries for one lifecycle.");
break;
}
clock.Advance(TimeSpan.FromSeconds(151));
EphemeralStoreSnapshot snapshot = store.GetSnapshot();
if (snapshot.ActiveListings != 0
|| snapshot.ActiveJoinAttempts != 0
|| snapshot.ReplayMarkers != 0
|| snapshot.IdempotencyEntries != 0
|| snapshot.RetainedOutcomeReports != 0)
{
failures.Add($"Accelerated soak retained state after cycle {cycle}.");
break;
}
cycle++;
}
elapsed.Stop();
GC.Collect();
GC.WaitForPendingFinalizers();
GC.Collect();
long managedGrowth = GC.GetTotalMemory(forceFullCollection: true) - managedBefore;
int handleGrowth = Process.GetCurrentProcess().HandleCount - handlesBefore;
if (managedGrowth > 67_108_864)
{
failures.Add($"Soak retained {managedGrowth} managed bytes; budget is 64 MiB.");
}
if (handleGrowth > 8)
{
failures.Add($"Soak retained {handleGrowth} process handles; budget is 8.");
}
return new(cycle, elapsed.Elapsed, peakScheduledExpiryEntries, managedGrowth, handleGrowth);
}
private static ObservedEndpoint PublicEndpoint(int port) =>
new(AddressFamilyKind.Ipv4, "203.0.113.10", port);
private static T RequireSuccess<T>(StoreResult<T> result, string operation)
{
if (!result.Succeeded)
{
throw new InvalidOperationException($"{operation} failed with {result.Code}.");
}
return result.Value!;
}
private static string ReadCpuModel()
{
const string cpuInfoPath = "/proc/cpuinfo";
if (!File.Exists(cpuInfoPath))
{
return "unavailable";
}
string? model = File.ReadLines(cpuInfoPath)
.FirstOrDefault(static line => line.StartsWith("model name", StringComparison.Ordinal));
int separator = model?.IndexOf(':') ?? -1;
return separator >= 0 ? model![(separator + 1)..].Trim() : "unavailable";
}
private static string ReadCgroupValue(string path) =>
File.Exists(path) ? File.ReadAllText(path).Trim() : "not-enforced";
private sealed class ManualClock : IWallClock, IMonotonicClock
{
public DateTimeOffset UtcNow { get; private set; } = DateTimeOffset.UtcNow;
public TimeSpan Elapsed { get; private set; }
public void Advance(TimeSpan duration)
{
UtcNow += duration;
Elapsed += duration;
}
}
private readonly record struct SoakEvidence(
int Cycles,
TimeSpan Duration,
int PeakScheduledExpiryEntries,
long ManagedGrowthBytes,
int HandleGrowth);
}
@@ -0,0 +1,13 @@
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<OutputType>Exe</OutputType>
<TargetFramework>net10.0</TargetFramework>
<AssemblyName>FinalFactory.Rendezvous.Capacity</AssemblyName>
<RootNamespace>FinalFactory.Rendezvous.Capacity</RootNamespace>
<IsPackable>false</IsPackable>
</PropertyGroup>
<ItemGroup>
<ProjectReference Include="../../src/FinalFactory.Rendezvous.Contracts/FinalFactory.Rendezvous.Contracts.csproj" />
<ProjectReference Include="../../src/FinalFactory.Rendezvous.Server/FinalFactory.Rendezvous.Server.csproj" />
</ItemGroup>
</Project>
@@ -0,0 +1,37 @@
using System.Text.Json;
namespace FinalFactory.Rendezvous.Capacity;
internal static class Program
{
private static readonly JsonSerializerOptions JsonOptions = new(JsonSerializerDefaults.Web)
{
WriteIndented = true,
};
public static async Task<int> Main(string[] args)
{
CapacityOptions options;
try
{
options = CapacityOptions.Parse(args);
}
catch (ArgumentException exception)
{
Console.Error.WriteLine(exception.Message);
return 2;
}
CapacityReport report = await CapacityRunner.RunAsync(options).ConfigureAwait(false);
string json = JsonSerializer.Serialize(report, JsonOptions);
Console.WriteLine(json);
if (options.OutputPath is not null)
{
string fullPath = Path.GetFullPath(options.OutputPath);
Directory.CreateDirectory(Path.GetDirectoryName(fullPath)!);
await File.WriteAllTextAsync(fullPath, json + Environment.NewLine).ConfigureAwait(false);
}
return report.Passed ? 0 : 1;
}
}

Some files were not shown because too many files have changed in this diff Show More