Prevents OOM when switching between models — only one 14B model
fits in 16GB VRAM at a time with KV cache for context.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Update image tag to v2026.2.0-efa9806-nix — the first source-built
authentik container from the build-authentik-from-source chain.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace floating :18 tag with pinned :18.3 (upstream out-of-cycle
release fixing 18.2 regressions). Stamps service as reviewed.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The external-secrets webhook injects conversionStrategy, decodingStrategy,
and metadataPolicy defaults on admission. Declaring them explicitly prevents
ArgoCD SSA from flagging the resource as OutOfSync.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
- Add `cluster` label (indri/ringtail) to all Prometheus scrape jobs, Alloy k8s metrics/logs, and Alloy host metrics/logs
- Deploy kube-state-metrics on ringtail's k3s cluster (ArgoCD app + manifests)
- Deploy Alloy on ringtail to collect pod metrics and logs, remote-writing to indri's Prometheus and Loki
- Replace single-cluster "Minikube Kubernetes" and "K8s Services Health" dashboards with:
- **Kubernetes Clusters** dashboard — multi-cluster with `cluster` and `namespace` template variables
- **Ringtail (k3s)** dashboard — dedicated ringtail view with GPU usage panels
## Deployment and Testing
1. Sync `apps` on indri ArgoCD to pick up new app definitions (`kube-state-metrics-ringtail`, `alloy-ringtail`)
2. Sync `prometheus` → verify `cluster` label on scraped metrics
3. Sync `alloy-k8s` → verify `cluster=indri` on remote-written metrics and logs
4. Run `mise run provision-indri -- --tags alloy` → verify `cluster=indri` on host Alloy metrics/logs
5. Sync `kube-state-metrics-ringtail` → verify pods running on ringtail
6. Sync `alloy-ringtail` → verify pods running, check Prometheus for `kube_pod_info{cluster="ringtail"}`
7. Sync `grafana-config` → verify dashboards appear, cluster variable populates both values
8. Check Loki for `{cluster="ringtail"}` logs from ringtail pods
## Notes
- Alloy on ringtail uses `insecure_skip_verify=true` for TLS to Prometheus/Loki (Tailscale-managed certs not in container trust store) — tighten later
- DNS resolution for `*.tail8d86e.ts.net` from ringtail pods depends on CoreDNS inheriting host's MagicDNS resolver; may need CoreDNS forwarding rules if pods can't resolve
- The old services dashboard (blackbox probes) is removed — those probes are still running in alloy-k8s and the data is still in Prometheus, just not in a dedicated dashboard
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/270
Parked car was being re-detected every few minutes at night due to IR
illumination noise triggering motion detection. Restrict the driveway
zone to [person, dog, cat] so cars and birds no longer create events
there. Cars still alert via the driveway_entrance zone.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
- Point ArgoCD app directly at forge-mirrored upstream repo (`mirrors/cloudnative-pg`) instead of the Helm charts repo
- Use `directory.include` to select the specific release manifest (`cnpg-1.27.1.yaml`) from the `releases/` directory
- No vendored files, no Helm — upgrades are a two-line change (`targetRevision` + `directory.include`)
- Delete unused `values.yaml` (was empty, all Helm defaults)
## Deployment and Testing
- [ ] Register mirror repo in ArgoCD: `argocd repo add ssh://forgejo@forge.ops.eblu.me:2222/mirrors/cloudnative-pg.git --ssh-private-key-path <key>`
- [ ] `argocd app set cloudnative-pg --revision feature/cnpg-direct-source && argocd app sync cloudnative-pg`
- [ ] Verify operator pod running: `kubectl get pods -n cnpg-system --context=minikube-indri`
- [ ] Verify CRDs exist: `kubectl get crd --context=minikube-indri | grep cnpg`
- [ ] Verify existing clusters healthy: `kubectl get clusters -A --context=minikube-indri`
- [ ] After merge: `argocd app set cloudnative-pg --revision main && argocd app sync cloudnative-pg`
## Notes
- The forge mirror was created via `mise run mirror-create` from `https://github.com/cloudnative-pg/cloudnative-pg.git`
- ArgoCD may need the mirror repo added to its known repositories if the credential template doesn't already match `mirrors/*`
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/268
The panel queried frigate_camera_events but the actual metric exposed
by Frigate is frigate_camera_events_total with a "camera" label
(not "camera_name").
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
- Widen `repo-creds-forge` URL prefix from `/eblume/` to host-wide `/` so it matches repos in all forge orgs (fixes `mirrors/` repos not getting SSH credentials)
- Update 8 ArgoCD app definitions from `eblume/<mirror>` → `mirrors/<mirror>` (immich-charts, cloudnative-pg-charts, external-secrets, connect-helm-charts)
- Fix stale alloy clone comment in Ansible defaults
- Bump immich v2.5.2 → v2.5.6 (bug-fix patches only)
- Update ArgoCD README bootstrap command and credential docs
## Context
Mirrors were migrated from `forge.ops.eblu.me/eblume/` to `forge.ops.eblu.me/mirrors/` in commit `cd57814`. Container Dockerfiles and image tags were updated, but ArgoCD app definitions and the repo credential template were missed, causing `ComparisonError` on apps that source Helm charts from mirrored repos.
## Deployment
1. Sync the ArgoCD `argocd` app first (picks up the widened credential template)
2. Sync the `apps` app (picks up new repo URLs for all 8 apps)
3. Verify immich resolves its ComparisonError: `argocd app get immich`
4. Sync immich to deploy v2.5.6: `argocd app sync immich`
5. Spot-check: `argocd app get external-secrets`, `argocd app get cloudnative-pg`, `argocd app get 1password-connect`
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/266
Prometheus and teslamate builds still in progress — will update in a
follow-up commit once their 33b7f0f tags land.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The hand-written configmap.yaml had app.kubernetes.io/name and
app.kubernetes.io/instance labels; configMapGenerator dropped them.
Add options.labels to both generator entries to restore parity.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
- Move hardcoded image tags to kustomization.yaml `images:` transformer across **22 services** — image names in manifests become version-agnostic templates, with tags centralized in one place per service
- Replace hand-written ConfigMap manifests with `configMapGenerator:` in **12 services** — config data extracted to standalone files, generated ConfigMaps include content hashes that trigger automatic pod rollouts on changes
- Create new `kustomization.yaml` for **forgejo-runner** and **nvidia-device-plugin** (switches ArgoCD from directory mode to kustomize mode, rendered output identical)
### Services modified
**Images only (8):** cv, devpi, docs, kube-state-metrics, miniflux, navidrome, teslamate, torrent
**Images + configMapGenerator (10):** alloy-k8s, forgejo-runner, frigate, grafana, homepage, kiwix, loki, mosquitto, ntfy, prometheus
**Images only, no configMapGenerator (4):** authentik (skip blueprints — special YAML tags), tailscale-operator-base (Deployment only, CRD image fields left as-is)
**Skipped entirely (6):** argocd (remote upstream), databases (no image fields), external-secrets, grafana-config (cross-kustomization dashboards), immich (Helm-managed), 1password-connect/cloudnative-pg (no kustomization.yaml)
### What changes at deploy time
- **images:** — no functional diff, `kustomize build` produces identical output with tags
- **configMapGenerator:** — ConfigMap names gain hash suffixes (e.g., `prometheus-config` → `prometheus-config-6f42fhctcb`) and all Deployment/StatefulSet/DaemonSet references are updated automatically. Pods will restart once per service on first sync due to the name change
## Test plan
- [x] `kubectl kustomize` builds all 30 service directories successfully
- [x] Image tags verified in rendered output for all modified services
- [x] ConfigMap hash suffixes verified in rendered output
- [x] ConfigMap references in Deployments/StatefulSets confirmed to use hashed names
- [x] All pre-commit hooks pass (yamllint, shellcheck, prettier, etc.)
- [ ] `argocd app diff` each service to confirm only expected ConfigMap name changes
- [ ] Deploy from branch starting with a low-risk service (e.g., mosquitto)
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/264
Grafana 12.x's grafana-postgresql-datasource plugin requires the
database name in jsonData, not just the top-level database field.
Without it, the frontend blocks all queries with "no default database
configured", causing all TeslaMate panels to show "No Data."
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The INI parser was stripping outer single quotes from
role_attribute_path = 'Admin', causing Grafana to evaluate 'Admin'
as a JMESPath field identifier instead of a string literal. This
resulted in all OAuth users getting the default Viewer role.
Replaced with a proper group-based expression that checks for the
'admins' Authentik group and maps to Admin/Viewer accordingly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Prometheus: v3.9.1-74029e1 [branch] -> v3.9.1-2ba5d8a [main]
Grafana: v12.3.3-09ac36b [branch] -> v12.3.3-d05d2fb [main]
These images were built during PR development and referenced branch
commits that won't survive branch cleanup. The [main] tags are
identical rebuilds from the squash-merge commit.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
- `foldersFromFilesStructure` was `false` in Grafana's sidecar provider config, causing Grafana to ignore the subdirectory structure the sidecar creates from `grafana_folder` annotations
- All 18 TeslaMate dashboards were appearing in the root "Dashboards" folder despite having `grafana_folder: "TeslaMate"` annotations on their ConfigMaps
- Flipping to `true` makes Grafana replicate the sidecar's directory structure as UI folders
## Deployment and Testing
- [ ] Sync `grafana` app: `argocd app sync grafana`
- [ ] Verify TeslaMate dashboards appear under a "TeslaMate" folder in Grafana's dashboard list
- [ ] Verify other dashboards remain in the root "Dashboards" folder
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/253
## Summary
Completes the `upgrade-k8s-runner` mikado chain. Both prerequisites (workflow validation in Dagger, config review against v12 defaults) were resolved in #250.
- Bump runner image `code.forgejo.org/forgejo/runner:6.3.1` → `12.7.0`
- Update `service-versions.yaml` to track new version
- Mark goal card complete (remove `status: active`)
## Deployment and Testing
After merge:
1. `argocd app sync forgejo-runner`
2. Verify runner registers in Forgejo admin → runners
3. Trigger a test workflow (e.g. `branch-cleanup.yaml` manual dispatch)
Rollback: revert image tag to `6.3.1`, push, sync.
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/251
## Summary
- Review runner config against v12.7.0 defaults — added `shutdown_timeout: 3h`, no breaking changes found
- Add `validate_workflows` Dagger function using `forgejo-runner validate --directory .` inside upstream container
- All 6 workflows pass v12.7.0 schema validation
- Wire `mise run validate-workflows` task and pre-commit hook on `.forgejo/workflows/` changes
- Mark both leaf Mikado cards (`review-runner-config-v12`, `validate-workflows-against-v12`) complete
## Mikado State
After merge, `upgrade-k8s-runner` goal card has no unmet dependencies — ready to execute the actual image bump in a follow-up PR.
## Test Plan
- [x] `dagger call validate-workflows --src=.` passes (all 6 workflows OK)
- [x] Pre-commit hooks pass
- [ ] Reviewer: confirm `shutdown_timeout: 3h` addition to ConfigMap looks reasonable
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/250
## Summary
- Switch from MQTT to webapi polling (v0.5.4 requires only one method)
- Poll every 15s for responsive alerts
- **`notify_once: true`** — one notification per event instead of repeats as object changes zones
- **`nosnap: drop`** — skip events without snapshots (was causing all events to be dropped on v0.3.5)
- **`snap_hires: true`** — use recording stream for higher quality snapshot images
## Deployment and Testing
- [ ] Sync: `argocd app set frigate --revision fix/frigate-notify-config && argocd app sync frigate`
- [ ] Verify pod starts: `kubectl --context=k3s-ringtail -n frigate get pods -l app=frigate-notify`
- [ ] Check logs for successful startup and event processing (no "No snapshot" drops)
- [ ] Wait for a motion event and confirm single ntfy notification with hi-res snapshot
- [ ] After merge: `argocd app set frigate --revision main && argocd app sync frigate`
Reviewed-on: https://forge.ops.eblu.me/eblume/blumeops/pulls/242
Five container manifests were removed when deleting old-style tags
(shared digests). Rebuild on a72a0d8 and update references.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
