Commit Graph

5 Commits

Author SHA1 Message Date
chrisfu
15890187a3 Stabilize CNPG reset/update flow and finalize 3-node recovery
- add idempotent CNPG operator reconciliation to prevent duplicate active controller ReplicaSets
- restore and validate CNPG 3-instance convergence with PV claimRef recovery and storage path readiness
- wire canonical launcher aliases/reset behavior and improve namespace cleanup semantics
- harden backup/objectstore readiness handling and retry behavior
- enforce service namespace usage for common services and remove default-namespace drift
- enable Kerberos milestone auto-activation when realm/kdc are configured and apply CNPG GSS pg_hba rules
- keep final runtime healthy: knoe-db 3/3, operator stable, perfsnap captured

Co-authored-by: Junie <junie@jetbrains.com>
2026-03-27 21:54:53 -07:00
chrisfu
a43aed7134 k3s: eliminate localhost registry + dedupe common-core
- Ensure k3s mode uses the k3s registry endpoint and avoid localhost/k3d image prefixes.

- Make ArgoCD repo-server cmp symlink creation idempotent.

- Normalize common-core provisioning to knoe-system and add repair-time dedupe of stray default-namespace installs.

- Add k3s MariaDB datastore/refresh playbooks and regression tests.
2026-03-05 14:31:33 -08:00
chrisfu
ec8a4e98af k3s: Fix installation hangs, CA mismatches, and arm64 networking
- Implement proactive CA hash verification and automated repair for mismatches

- Ensure agents prioritize discovered server tokens over stale vault values

- Fix K3s service hangs with explicit stop and killall before reinstall

- Add Retropie/Pi networking fixes (WiFi power save, wlan0 priority)

- Pin pre-staged images to stable, architecture-aware versions (arm64)

- Remove obsolete init-port-forwards and prole.cfg sync tasks

- Update k8s manifests and installer core logic with new tests
2026-02-16 23:30:27 -08:00
chrisfu
6b6e5e2aec nodeSelector and storage updates across k8s manifests
- Added `nodeSelector` for multiple Kubernetes resources to ensure scheduling on `myrddin.prole.org`.
- Modified `storage` requests and set `storageClassName` in `garage-statefulset.yaml`.
- Enhanced `install.py` for dynamic environment configuration and kubeconfig handling.
- Improved cgroup management tasks in Ansible with conflict resolution and parameter updates.
- Simplified vault token update process in playbooks and updated encryption checks.
2026-02-06 01:17:54 -08:00
chrisfu
da2f6600ba feat: infrastructure and installer updates for k3s, OpenTofu, and prole-db
- Add k3s start/stop Ansible playbooks and roles.

- Implement OpenTofu initialization scripts and k8s manifests.

- Update ncurses installer with OpenTofu support and improved k3s integration.

- Add mode support (--mode) to etc/ initialization scripts.

- Update prole-db with recovery, barman objectstore, and SSH OpenBao support.

- Refine k8s manifests for OpenBao and prole-db.
2026-02-05 21:27:18 -08:00