From cbf2a8ee39f3828edf3dbf42b751c9cc0f619d19 Mon Sep 17 00:00:00 2001 From: ekklesia Date: Mon, 5 Oct 2026 21:05:15 -0500 Subject: [PATCH] ci: the integration job removes its stacks even when the runner kills it The platform's rule (its ci-sweep.sh, docs/development/security.md), applied to the one Signage job that starts Docker stacks: - it has its own EKKLESIA_CI_JOB_KEY; - it sweeps stale CI stacks from the runner before it starts; - it removes its own stacks in a final `if: always()` step, which tolerates a failed assembly, since nothing was started then; - the image ci/integration.sh builds is named and labeled as a CI stack of this run and job, with the time CI built it, so the sweep finds it. ci/platform.env moves to the platform commit that adds the sweep (Ekklesia #60). The assembled lockfile is unchanged, and still passes a frozen install. Verified: SIGKILL of the integration script's whole process group mid-run left 11 resources (2 containers, 4 networks, 4 volumes and the image). `ci-sweep.sh --job` removed exactly those 11. A normal run passes, and the job sweep after it finds nothing. Co-Authored-By: Claude Opus 5.5 --- .gitea/workflows/ci.yml | 17 +++++++++++++++++ ci/integration.sh | 8 +++++++- ci/platform.env | 2 +- 3 files changed, 25 insertions(+), 2 deletions(-) diff --git a/.gitea/workflows/ci.yml b/.gitea/workflows/ci.yml index 5aac322..d8b50f0 100644 --- a/.gitea/workflows/ci.yml +++ b/.gitea/workflows/ci.yml @@ -187,6 +187,9 @@ jobs: timeout-minutes: 45 env: DOCKER_BUILDKIT: "1" + # Names this job's stacks, so its final step removes exactly these (the platform's + # scripts/lib/ci-project.sh). + EKKLESIA_CI_JOB_KEY: signage-integration steps: - name: Check out repository uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2 @@ -196,5 +199,19 @@ jobs: - name: Assemble Signage into the pinned platform run: bash ci/assemble.sh + # A job the runner killed (cancelled, or past the runner's own time limit) never ran its + # cleanup. Anything a CI job left on this shared daemon more than three hours ago goes first. + - name: Sweep stale CI stacks from the runner + run: bash build/workspace/scripts/ci-sweep.sh --stale + - name: Build the image and run the display scenarios run: bash ci/integration.sh + + # Whatever happened above, this job's own stacks do not outlive it. If assembly failed, nothing + # was started. + - name: Remove this job's CI stacks + if: always() + run: | + if [ -f build/workspace/scripts/ci-sweep.sh ]; then + bash build/workspace/scripts/ci-sweep.sh --job + fi diff --git a/ci/integration.sh b/ci/integration.sh index a22bd0c..b5ebaa4 100755 --- a/ci/integration.sh +++ b/ci/integration.sh @@ -15,7 +15,12 @@ workspace="$root/build/workspace" exit 1 } -image="signage-ci:${GITHUB_SHA:-local}-${GITHUB_RUN_ID:-0}-${GITHUB_RUN_ATTEMPT:-1}" +# Named and labeled as a CI stack of this run and job (the platform's scripts/lib/ci-project.sh), so +# the workflow's final `ci-sweep.sh --job` removes it even if this script is killed before its trap. +# shellcheck source=/dev/null +source "$workspace/scripts/lib/ci-project.sh" +project="$(ci_project_name signage-image)" +image="${project}:latest" trap 'docker image rm --force "$image" >/dev/null 2>&1 || true' EXIT scenarios="$(sed -e 's/#.*//' -e '/^[[:space:]]*$/d' "$root/ci/display-scenarios" | tr '\n' ' ')" @@ -25,6 +30,7 @@ scenarios="$(sed -e 's/#.*//' -e '/^[[:space:]]*$/d' "$root/ci/display-scenarios } DOCKER_BUILDKIT=1 docker build --quiet --file "$workspace/deploy/Dockerfile" --tag "$image" \ + --label "${EKKLESIA_CI_LABEL}=${project}" --label "${EKKLESIA_CI_CREATED_LABEL}=$(date +%s)" \ "$workspace" >/dev/null echo "integration: built the deployment image with Signage; scenarios: ${scenarios}" diff --git a/ci/platform.env b/ci/platform.env index de7e07f..41cf4b3 100644 --- a/ci/platform.env +++ b/ci/platform.env @@ -3,4 +3,4 @@ # # Move it deliberately: change the commit, run `bash ci/update-lock.sh`, and commit both. EKKLESIA_PLATFORM_REPOSITORY=https://git.riverofglorymn.org/River_of_Glory_Church/Ekklesia.git -EKKLESIA_PLATFORM_COMMIT=f50e8f42018c15698709739e2ba54771e97df08a +EKKLESIA_PLATFORM_COMMIT=00e544125b855372e0f173db33323a57ac357d02