STABLE BUILD — all 35+ pods Running, zero non-Running pods, CNPG ready=3.
## Architecture changes (Supabase slim-down)
- Removed local Supabase Kong, Storage, MinIO, Imgproxy from deployed stack
- Rewired API routing to shared Kong in kube-system
- Rewired object storage to shared Garage in knoe-system
- Supabase footprint reduced to control-plane only: auth, rest, meta,
analytics, realtime (studio port-forward retained)
- Supabase pods renamed: stripped redundant '-prole-supabase-' prefix via
per-component fullnameOverride (e.g. supabase-auth, supabase-rest, ...)
- Supabase workloads pinned to merlin.prole.org (memory-safe node)
## File changes
- conf/service/prole.cfg
* Replaced all hardcoded /Users/chrisfu/dev/prole paths with ${HOME}/dev/prole
so config is portable when run from myrddin.prole.org or any user home
* SUPABASE_ENABLED=True, SUPABASE_PRIMARY_NODE=merlin.prole.org in
both [Global] and [Optional Features]
* supabase_enabled=true in [Inputs]/init_cluster section
- supabase/helm/render_supabase.py
* Explicitly disabled: kong, storage, minio, imgproxy
* Active components pinned via nodeSelector to merlin.prole.org
* fullnameOverride per component to clean up pod names
* Default primary node changed from pi.prole.org -> merlin.prole.org
- supabase/deploy.sh
* Removed svc/kong port-forward; API via shared Kong in kube-system
* Removed imgproxy/storage from PV/PVC prep and helm --set persistence args
* Fixed kubectl exec missing -i flag for psql heredoc stdin in
setup_knoe_db_for_supabase
* Updated Studio access messaging to reference shared platform Kong
- k8s/prole/knoe-db.yaml
* instances: 3
* Added control-plane/master tolerations for myrddin scheduling
- deploy/opentofu/k3s/manifests/prole/knoe-db.yaml
* instances: 3, added control-plane tolerations (mirrors k8s/prole/)
- k8s/prole/iscsi-pvs.yaml
* Expanded d003-data/d003-wal nodeAffinity to [pi, merlin, myrddin]
to reflect post-OOM-recovery live state
- k8s/registry/deployment.yaml
* nodeSelector restored to myrddin.prole.org (hostPort 5000 owner)
* Added control-plane + master tolerations so pod schedules correctly
- modes/k3s/knoe-db/.version
* Aligned to 140 (image tag 18-140) matching repo-root version
- etc/init_common_services.sh, etc/init_registry.sh
* Minor fixes aligned with slimmed service layout and node targeting
- infrastructure/inventory/host_vars/pi.prole.org.yml
* Updated host vars to reflect post-recovery pi state
- knoe/core/topology.py, knoe/ui/screens/cluster.py,
knoe/ui/screens/cluster_nodes.py
* Topology and UI updates for 3-node cluster representation
- scan/network_description.txt
* Refreshed network scan output post-recovery
- conf/port-mapping.cfg
* Updated port mapping to reflect removal of local Kong/storage forwards
- scripts/cleanup_stale_rs.sh (new)
* Utility script to prune orphaned ReplicaSets after rollouts
## Verified stable state
- CNPG: Cluster in healthy state, ready=3
knoe-db-2 (merlin, primary), knoe-db-3 (pi), knoe-db-5 (myrddin)
- Supabase: 5 pods Running on merlin.prole.org
supabase-auth, supabase-rest, supabase-meta,
supabase-analytics, supabase-realtime
- Monitoring: all prometheus/grafana/alertmanager pods Running on merlin
- Registry: Running on myrddin.prole.org (hostPort 5000)
- Zero non-Running / non-Completed pods across all namespaces
Co-authored-by: Junie <junie@jetbrains.com>
- Rename host mount namespace from /prole/d00x to /synology/d00x
- Make LUN ownership explicit per host in inventory (myrddin=d001; merlin=d002,d004; pi=d003)
- Add safe migration cleanup for legacy /prole/d00x mounts and stale /etc/fstab entries
- Update k3s/ArgoCD/OpenTofu manifests and PV node affinities to match new mount paths
- Improve iSCSI role check-mode behavior and add quiesce/detach flow in ansible.sh
- Pin Prometheus/Alertmanager + Grafana to merlin.prole.org in k3s to avoid PV node-affinity issues
- Taint myrddin.prole.org as control-plane-only and add CNPG toleration
- Prevent init_cloudnative_pg from deploying prole-svc-kong into the DB namespace
- GitOps-manage Kong declarative config via ConfigMap + add placement/init tests
Prefer infra-managed manifests for bootstrap; reconcile CNPG instances based on Ready+schedulable labeled db nodes; update manifests to use node-role affinity + anti-affinity; add policy/tests and config touch-ups (incl. prole.cfg).
Update inventory/roles to keep K3s state on local storage and prevent iSCSI from managing /var/lib/rancher.
Also add a single-host k3s install playbook, a systemd override template, and docs describing the installer ↔ Ansible boundary and slow-storage knobs.
Co-authored-by: Junie <junie@jetbrains.com>
- k3s_sync: choose any server as sync source fallback; refresh kubeconfig and validate nodes\n- fetch_kubeconfig: save to prole-k3s.kubeconfig and etc/secrets/k3s.kubeconfig\n- k3s sync: sync server URL into config.yaml; restore TLS bundle only on servers\n- cert-manager/ACME: wait for CNI/webhook readiness; retry ClusterIssuer/Certificate apply\n- inventory: tune myrddin control-plane timing args; add render test\n- misc: update port mappings and prole-db manifests/version
Co-authored-by: Junie <junie@jetbrains.com>
- Ensure dashboard service doesn't create an ordering cycle with `multi-user.target`.
- Verify service safely attaches to a virtual terminal (VT) and passes VT to Xorg.
- Ensure no `su` usage by default in the service.
- Validate script behavior for passing VT and `-keeptty` options to Xorg.
chore(dashboard): update systemd unit to avoid VT conflicts and improve logging
- Remove `After` and `Wants` directives for `multi-user.target` to prevent conflicts.
- Add `TTYPath`, `StandardInput`, and related options for VT attachment.
- Ensure `ExecStopPost` restores getty service for the dashboard VT.
fix(dashboard): improve Xorg resolution and clean up script logic
- Dynamically determine Xorg executable location in `dashboard.sh`.
- Pass VT, display, and `-keeptty` to Xorg to improve script reliability.
fix(mariadb): correct exFAT mount options for external storage
- Avoid `chown` failures by setting root ownership with group access for `mysql`.
- Update external mount options to `uid=0,gid=mysql,umask=0002`.
cleanup(mariadb): remove unnecessary owner/group permissions from migration marker task
- Enable mounting borrowed USB disk (exFAT) persistently at `/external`.
- Migrate MariaDB datadir to `/external/mariadb` with bind mount at `/srv/mariadb/mariadb`.
- Add migration marker and rollback support to prevent accidental remigration.
- Include Ansible tasks for external storage setup and migration validation.
- Added tasks and playbooks for K3s datastore export/import using MariaDB Tools role, with associated tests and defaults.
- Enhanced iSCSI role to support mkfs-once logic and safer re-initialization of block storage.
- Migrated iSCSI-backed Rancher data from retropie to merlin.prole.org.
- Updated k3s roles/playbooks to relax Rancher storage preflight checks, supporting PARTUUID-based mounts.
- Adjusted Samba AD NetBIOS name derivation to use uppercase short hostname by default.
- Incremented prole DB version to 104, updated generated prole.cfg, inventory, and recovery manifest templates.
ArgoCD IO load distribution: Updated k8s/argocd/install.yaml to use hostPath volumes for application-controller (/prole/d001), redis (/prole/d002), and repo-server (/prole/d003). Updated Ansible configure task to create hostPath directories on myrddin.prole.org.
K3s configuration and testing: Added test playbooks for kubeconfig rewriting and argument validation. Updated K3s install, configure, and image import tasks. Configured required mounts and host-specific variables for myrddin.prole.org.
Miscellaneous: Updated conf/prole.cfg with K3S_TOKEN. Bumped prole-db version and updated recovery manifest templates. Minor UI adjustments in cluster screen.
Co-authored-by: Junie <junie@jetbrains.com>
- Fixed YAML syntax and Python version parsing in CNPG operator tasks.
- Added automated detection of active network interface (eth0 vs wlan0) for Retropie and Pi nodes to ensure correct K3s configuration.
- Implemented persistent node taints for retropie.prole.org to exclude CloudNative-PG workloads.
- Added a rollout restart mechanism for cnpg-controller-manager to recover from scheduling on tainted or unstable nodes.
- Removed redundant kubectl-cnpg dependency from the Python installer as it is now managed by Ansible.
- Improved prole.sh to force color output for Ansible playbooks.
- Updated prole.cfg with active K3s token and CNPG version information.
- 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
Repairs and improvements:
- iSCSI: Added cleanup tasks to remove stale mounts and fstab entries. Improved robustness of iSCSI target management and added 'iscsi_absent_mounts' support.
- K3s:
- Updated service start logic to accept 'activating' state, preventing premature failure during slow startups.
- Improved service stop logic to safely handle missing or not-found services.
- Ensured 'prole-installer' ServiceAccount and ClusterRoleBinding exist for K8s administration.
- Added leader election and etcd tuning arguments (forgiving leases) to config.yaml.j2.
- Removed deprecated 'prole-port-forwards' systemd service.
- Installer & Scripts:
- Updated legacy_tk.py to support K3s mode, secret resolution for passwords, and better environment management (including ~/.prole/env.sh for service mode).
- Updated init_ansible.sh to support PROLE_VAULT_PASS_FILE and ANSIBLE_VAULT_PASSWORD_FILE.
- Improved directory and kubeconfig path resolution in prole_cfg.sh to support fallback to ~/.prole.
- Enhanced Grafana password resolution in init_monitoring.sh.
- Added automatic application of iSCSI StorageClass and PersistentVolumes in init_openbao.sh.
- General: Switched conf/prole.cfg to k3s deployment mode and updated vault_k3s.yml token.
New Ansible Tasks and Playbooks:
- infrastructure/playbooks/iscsi_cleanup.yml: Automates logout and removal of stale iSCSI node records.
- infrastructure/playbooks/prole_logs_migrate.yml: Orchestrates /prole/logs migration to iSCSI storage.
- infrastructure/playbooks/tmp_bao_dir.yml: Ensures host-level storage directories for OpenBao.
- infrastructure/playbooks/tmp_mount.yml: Utility to verify and enforce host-level mounts.
- infrastructure/playbooks/k3s_sync.yml: Added tasks to start K3s after sync and update local kubeconfig on the controller.
- Added 'Unmount stale iSCSI mounts' and 'Remove stale iSCSI fstab entries' to the iscsi role.
- Added 'Ensure prole-installer service account exists' to the k3s role.
- install.py: Major update including configuration variable expansion, improved k3s/k3d handling, and enhanced installation logic.
- etc/ scripts: Significant refactoring of initialization scripts (Kerberos, Port Forwards, Garage Store, etc.).
- Port Forwards: Transitioned from XML to port-mappings.conf for managing kubectl port-forwards.
- Status Reporting: Improved status checking for common services.
- Infrastructure: Updated Ansible inventory and rsyslog role configurations.
- Tests: Added a comprehensive suite of tests for 'etc' initialization scripts in prole/tests/etc/.
- Documentation: Added prole-db-documentation-mcp-architecture.md.
- General: Updated Dockerfiles and various helper scripts.
- Introduced cgroup kernel parameter management for `k3s_hosts` via Ansible playbook.
- Added Kubernetes PersistentVolumes for iSCSI storage with specific node affinity.
- Defined `prole-iscsi` StorageClass for volume provisioning.
