rock
ad1ce5e4aa
All CI workflows across all repos were broken. Three root causes:
1. Runner labels pointed to bare Alpine image (forgejo/runner:6) which has
no Node.js, no docker, no Go/Rust, no root, no apt-get. Every workflow
step failed immediately.
Fix: Change labels to official Debian language images:
golang → docker://golang:1.26-bookworm
node → docker://node:22-bookworm
rust → docker://rust:1-bookworm
2. Docker socket not shared between dind sidecar and runner container.
dind creates /run/docker.sock in its own filesystem. Runner container
has a separate filesystem and can't see it. Workflow containers that
run 'docker build' get 'Cannot connect to Docker daemon'.
Fix: Share /run between dind and runner via emptyDir volume.
Add docker_host: automount to runner config so the socket is
mounted into workflow containers automatically.
Note: /var/run is a symlink to /run in Alpine — must mount at /run.
3. Init container skipped re-registration if .runner file existed on PVC.
Changing labels in values.yaml had no effect until PVC was manually
deleted — not GitOps-friendly.
Fix: Always rm .runner and re-register on pod start. Labels stay
in sync with values.yaml automatically.
Also removes custom Dockerfiles and build-runner-images workflow — no longer
needed since official images provide the language tools directly.
60 lines
1.9 KiB
YAML
60 lines
1.9 KiB
YAML
runner:
|
|
image:
|
|
repository: code.forgejo.org/forgejo/runner
|
|
tag: "6"
|
|
name: golang-runner
|
|
# Label image is what workflow steps run in (NOT the runner daemon image).
|
|
# golang:1.26-bookworm: Debian, root, apt-get, Go, git.
|
|
# Install Node.js/docker in workflow steps as needed.
|
|
labels: "golang:docker://golang:1.26-bookworm"
|
|
forgejoUrl: http://forgejo-gitea-http.cicd.svc.cluster.local:3000
|
|
tokenSecret: runner-token
|
|
resources:
|
|
requests:
|
|
cpu: 100m
|
|
memory: 256Mi
|
|
limits:
|
|
cpu: "2"
|
|
memory: 4Gi
|
|
|
|
dind:
|
|
image:
|
|
repository: docker
|
|
tag: "27-dind" # pin exact release before apply
|
|
resources:
|
|
requests:
|
|
cpu: 100m
|
|
memory: 256Mi
|
|
limits:
|
|
cpu: "2"
|
|
memory: 4Gi
|
|
|
|
persistence:
|
|
reg:
|
|
storageClass: longhorn # Unified StorageClass (3 replicas)
|
|
size: 20Gi # .runner registration file + action tool cache + actcache artifacts
|
|
dind:
|
|
storageClass: longhorn # Unified StorageClass (3 replicas)
|
|
size: 30Gi # docker layer cache — keeps rebuilds fast across restarts
|
|
|
|
tolerations:
|
|
- key: node-role.kubernetes.io/control-plane
|
|
operator: Exists
|
|
effect: NoSchedule
|
|
|
|
# Pin to az-b (talos-cp-2) — more Longhorn storage than az-a (worker-1 over-provisioned).
|
|
# RWO PVCs will recreate on talos-cp-2 when nodeSelector changes.
|
|
nodeSelector:
|
|
topology.kubernetes.io/zone: az-b
|
|
|
|
# GC CronJob — prunes Docker images/volumes/build-cache and actcache across
|
|
# ALL forgejo-runner pods. Only enable in default values (golang instance);
|
|
# disable in per-runner overrides so it renders once.
|
|
gc:
|
|
enabled: true
|
|
schedule: "*/30 * * * *" # every 30 minutes
|
|
image: alpine/k8s:1.31.0
|
|
pruneAge: "30m" # Docker artifacts unused longer than this get pruned
|
|
pruneAgeHours: 0.5 # Same as pruneAge but numeric for date arithmetic in shell
|
|
actcacheMaxAgeDays: 1 # actcache files older than N days (aggressive for heavy Rust cargo builds)
|