`act_runner`'s default `container.force_pull: true` re-issues a
`docker pull` at the start of every job, adding 10-30 s of registry
round-trip per job even when every layer is already locally cached.
The job images are already pinned to specific tags (catthehacker
ubuntu :act-22.04 + :full-22.04), so the implicit pull is both pure
overhead and contradictory with the deliberate-upgrade policy
ADR-0015 spells out for the runner image itself.
Mount a shared `infra/runner-config.yaml` into all three runners
with `force_pull: false`, point each runner at it via the
`CONFIG_FILE` env var, and document the pre-pull procedure (one-shot
`docker pull` on the runner host, plus the upgrade playbook) in
infra/README.md "Job image pinning and pre-pull".
The pre-pull is also folded into the "First-time registration"
walkthrough so a fresh setup is correct end-to-end.
## Summary
`act_runner`'s built-in GitHub-Actions-cache server binds inside the runner container on the compose-defined `apf-portal-act-runners` bridge. Jobs spawned via the mounted Docker socket land on Docker's default `bridge` network and can't reach it. Every job opting into `cache: 'pnpm'` ate ~2 min `ETIMEDOUT` on restore + another ~2 min on save — across the 5 jobs, ~20 min wasted per CI run for zero cache hits.
Drop `cache: 'pnpm'` everywhere. `pnpm install --frozen-lockfile` is fast on the warm store inside the job container, so removing the cache layer is a net gain today.
The proper fix (cross-container networking / fixed-port cache binding) is documented in `infra/README.md` → "Cache server (deferred)" so it can be picked up as an isolated infra spike later.
## Test plan
- [ ] CI run on this PR: every job's `Set up Node.js` step finishes in seconds (no ETIMEDOUT warning).
- [ ] `Complete job` step also finishes promptly (no `reserveCache failed` warning).
- [ ] Total wall-clock time for the run should drop by ~15-20 min vs. previous runs.
---------
Co-authored-by: Julien Gautier <julien.gautier@apf.asso.fr>
Reviewed-on: #8
Set up the infrastructure-as-code recipe to bring up the three
self-hosted Gitea Actions runners required by ADR-0015. The compose
file launches three act_runner instances pinned to 0.2.13, registered
with the project's Gitea organisation, labelled self-hosted + on-prem
to match the runs-on selector in every job under .gitea/workflows/*.
Layout:
- infra/ new top-level folder for IaC
- infra/README.md explains the folder, registration
flow, security implications, future
placeholders (local/, prod/, runbooks/)
- infra/ci-runners.compose.yml three act_runner services, networked
together, persisting credentials to
./data/runner-N
- infra/.env.example GITEA_INSTANCE_URL +
GITEA_RUNNER_REGISTRATION_TOKEN; .env
itself stays git-ignored (root rule)
- infra/data/.gitignore tracks the dir, ignores runtime state
Security posture (documented in infra/README.md): mounting
/var/run/docker.sock gives the runner root-equivalent access to the
host Docker daemon. Mitigations rely on (a) repo-scope of the runner
in Gitea, (b) running the runner host outside the production trust
boundary, (c) no extra host filesystem mounts. Future hardening
(rootless Docker, DinD sidecar) is flagged as deferred.
The compose pins the runner image (0.2.13). Bumps go through a
dedicated chore(deps) PR per the convention; image upgrades roll one
runner at a time so CI is never starved (procedure documented in the
README).
Doc indexing (docs/README.md) deliberately not touched here to avoid
a conflict with the pending docs/architecture-diagrams branch which
also modifies that file. A small follow-up PR will add an index entry
once that branch is merged.
Julien Gautier
julien