Day-2 Operations¶
Flux Reconciliation¶
Force Flux to pull the latest changes from Git:
Check the status of all Flux resources:
Talos Operations¶
Update Node Configuration¶
# Regenerate configs from templates
task talos:generate-config
# Apply to a specific node
task talos:apply-node IP=10.10.10.10 MODE=auto
Upgrade Talos¶
- Update
talosVersionintalenv.yaml - Run:
Upgrade Kubernetes¶
- Update
kubernetesVersionintalenv.yaml - Run:
Application Management¶
Suspend an Application¶
Resume an Application¶
Force Redeploy¶
Roll Back a HelmRelease¶
# Check history
helm history <release-name> -n <namespace>
# Rollback
helm rollback <release-name> <revision> -n <namespace>
Renovate¶
Renovate runs on a weekend schedule and creates PRs for dependency updates:
- Auto-merge: GitHub Actions (minor/patch), Mise tools (minor/patch)
- Manual review: Helm charts, container images (major versions)
- Dashboard: Check the "Dependency Dashboard" issue in GitHub
Adding a New Application¶
Follow the checklist in order:
- Create directory:
kubernetes/apps/<namespace>/<app-name>/app/ - Create manifests:
helmrelease.yaml,ocirepository.yaml,kustomization.yaml - Create
ks.yamlFlux Kustomization - Update
kubernetes/apps/<namespace>/kustomization.yaml - Encrypt any secrets with SOPS
- Consider Kanidm SSO integration
- Add to Homepage dashboard
- Add VolSync backup config if stateful
- Add monitoring (ServiceMonitor/PodMonitor) if metrics are exposed
- Add Discord alerts if critical
- Add NFS-scaler if mounting NFS volumes
- Update CLAUDE.md with the new application