kblack0610/home-config
1
0
Fork 0
Code Issues Pull requests 2 Projects Releases Packages Wiki Activity Actions

feat(ci): Forgejo pool counts, cc:host_hot, api_ok in ci/pools #616

Open
kblack0610 wants to merge 1 commit from feat/ci-forgejo-pool-host-hot-api-ok into master
pull from: feat/ci-forgejo-pool-host-hot-api-ok
merge into: kblack0610:master
Conversation 0 Commits 1 Files changed 5 +156 -5
kblack0610 commented 2026-11-06 09:40:33 +00:00
Owner
Copy link

The board row asked for three things. The Forgejo runner rows already landed in #520, so this PR covers what was still missing.

  • Forgejo pool counts. The exporter now emits github_runner_pool_expected / _online for pool="forgejo" as one host="k3s" row. Expected is the runner Deployment's replicas x the ConfigMap's capacity. A runner that is not ready, or a pod stuck Pending, now turns ci/pools warn and counts in cc:ci_online / cc:ci_expected. runner-reader-rbac.yaml gets get on that one Deployment.
  • api_ok in ci/pools. When a runner source fails, pool_online is absent and the short-pool count came back empty, so ci/pools read ok while blind. It now goes to 1 when any api_ok{kind="runners"} is 0, or when no runner source reports at all.
  • cc:host_hot{host}. 0 ok, 1 over 90C for 10m, 2 over 95C for 2m. Those are the NodeCpuHot and NodeCpuCritical thresholds and hold times, so the board marks the same machine the alert names. A new machines/thermal component reads it.

assumed: Forgejo pool is one host="k3s" row, because the pods float between nodes and no node has an expected size.
assumed: the workstation is left out of cc:host_hot, because its 7900X runs at 95C by design and the fleet_thermal watch already judges it.
assumed: no cc:red_now row for a hot host, because NodeCpuCritical already pages it.

Checks run locally:

  • test_exporter.py: 63 tests OK, 1 of them new, plus pool-row assertions added to the unreadable-runners test.
  • scripts/test-prometheus-rules.sh in alpine with promtool: all suites SUCCESS, with 4 new command-center cases. Negative control: with the api_ok clauses removed and the 90C line moved, 3 expectations failed. Then I restored both.
  • check-fleet-roster, check-runner-roster, test_governor, check-ha-packages, check-secret-store: all ok.

Not verified: live values after Flux applies (github_runner_pool_online{pool="forgejo"}, and cc:ci_expected going up by 4).

Item: ar-190
Task: ar-190

https://claude.ai/code/session_01AWkPssJ1GuZfnqeQ8Epbvj

The board row asked for three things. The Forgejo runner rows already landed in #520, so this PR covers what was still missing. - **Forgejo pool counts.** The exporter now emits `github_runner_pool_expected` / `_online` for `pool="forgejo"` as one `host="k3s"` row. Expected is the runner Deployment's replicas x the ConfigMap's capacity. A runner that is not ready, or a pod stuck Pending, now turns ci/pools warn and counts in `cc:ci_online` / `cc:ci_expected`. `runner-reader-rbac.yaml` gets `get` on that one Deployment. - **api_ok in ci/pools.** When a runner source fails, `pool_online` is absent and the short-pool count came back empty, so ci/pools read ok while blind. It now goes to 1 when any `api_ok{kind="runners"}` is 0, or when no runner source reports at all. - **cc:host_hot{host}.** 0 ok, 1 over 90C for 10m, 2 over 95C for 2m. Those are the NodeCpuHot and NodeCpuCritical thresholds and hold times, so the board marks the same machine the alert names. A new `machines/thermal` component reads it. assumed: Forgejo pool is one `host="k3s"` row, because the pods float between nodes and no node has an expected size. assumed: the workstation is left out of cc:host_hot, because its 7900X runs at 95C by design and the fleet_thermal watch already judges it. assumed: no `cc:red_now` row for a hot host, because NodeCpuCritical already pages it. Checks run locally: - `test_exporter.py`: 63 tests OK, 1 of them new, plus pool-row assertions added to the unreadable-runners test. - `scripts/test-prometheus-rules.sh` in alpine with promtool: all suites SUCCESS, with 4 new command-center cases. Negative control: with the api_ok clauses removed and the 90C line moved, 3 expectations failed. Then I restored both. - check-fleet-roster, check-runner-roster, test_governor, check-ha-packages, check-secret-store: all ok. Not verified: live values after Flux applies (`github_runner_pool_online{pool="forgejo"}`, and `cc:ci_expected` going up by 4). Item: ar-190 Task: ar-190 https://claude.ai/code/session_01AWkPssJ1GuZfnqeQ8Epbvj
feat(ci): Forgejo pool counts, cc:host_hot, api_ok in ci/pools
All checks were successful
checks / zomboid-scripts (pull_request) Successful in 5s
Details
checks / ci-disk-prune (pull_request) Successful in 5s
Details
checks / fleet-roster (pull_request) Successful in 8s
Details
checks / kustomize-builds (pull_request) Successful in 4s
Details
checks / prometheus-rules (pull_request) Successful in 8s
Details
2a5d25a138 
- exporter: github_runner_pool_expected/online for pool="forgejo" (host="k3s"), expected = runner Deployment replicas x capacity; RBAC to get that Deployment.
- ci/pools warns when a runner source reads api_ok=0 or none reports: pool_online is absent then, so it read ok blind.
- cc:host_hot{host} 0/1/2 on NodeCpuHot/NodeCpuCritical lines and holds, workstation left to fleet_thermal; machines/thermal component reads it.

Item: ar-190
Task: ar-190

Claude-Session: https://claude.ai/code/session_01AWkPssJ1GuZfnqeQ8Epbvj
All checks were successful
checks / zomboid-scripts (pull_request) Successful in 5s
Details
checks / ci-disk-prune (pull_request) Successful in 5s
Details
checks / fleet-roster (pull_request) Successful in 8s
Details
checks / kustomize-builds (pull_request) Successful in 4s
Details
checks / prometheus-rules (pull_request) Successful in 8s
Details
This pull request can be merged automatically.
This branch is out-of-date with the base branch
You are not authorized to merge this pull request.
View command line instructions

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u origin feat/ci-forgejo-pool-host-hot-api-ok:feat/ci-forgejo-pool-host-hot-api-ok
git switch feat/ci-forgejo-pool-host-hot-api-ok

Merge

Merge the changes and update on Forgejo.

Warning: The "Autodetect manual merge" setting is not enabled for this repository, you will have to mark this pull request as manually merged afterwards.

git switch master
git merge --no-ff feat/ci-forgejo-pool-host-hot-api-ok
git switch feat/ci-forgejo-pool-host-hot-api-ok
git rebase master
git switch master
git merge --ff-only feat/ci-forgejo-pool-host-hot-api-ok
git switch feat/ci-forgejo-pool-host-hot-api-ok
git rebase master
git switch master
git merge --no-ff feat/ci-forgejo-pool-host-hot-api-ok
git switch master
git merge --squash feat/ci-forgejo-pool-host-hot-api-ok
git switch master
git merge --ff-only feat/ci-forgejo-pool-host-hot-api-ok
git switch master
git merge feat/ci-forgejo-pool-host-hot-api-ok
git push origin master
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
kblack0610/home-config!616
No description provided.