Single-node Vault (raft storage, standalone mode) plus a restore CronJob
that pulls the latest raft snapshot from the VPS's own local MinIO
(already receiving snapshots every 6h from home) and restores it every
6h, offset 30 minutes after the home-side snapshot job.
Not an independent root of trust: raft snapshot restore replaces the
entire storage backend including the keyring, so after every restore
this Vault is sealed with HOME's actual keyring - unsealing it needs a
copy of home's real unseal key, stored directly on the VPS cluster
(kubectl, not git, same pattern as vps-minio-root-secret). First-run
bootstrap uses a throwaway single-Shamir-key init just to get through
the very first restore, then is irrelevant from then on.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Foundation for a DR/backup path using an always-on VPS as a second
ArgoCD-managed cluster, plus DB/backup standardization work that fell
out of it:
- vps-standby ArgoCD cluster destination + AppProject, MinIO backup
receiver, VPS bootstrap script (k3s, Netbird, cert-manager)
- Dual-site DNS failover watcher + home-IP DDNS CronJob, Cloudflare
token moved out of git into Vault+ExternalSecret
- Nextcloud migrated from ad-hoc MariaDB to CNPG + redis-operator
(matches n8n/Authentik/GitLab's backup-native pattern)
- Authentik's CNPG manifests moved into the actual ArgoCD-synced
manifests/ path (were present but never wired into the sync path)
- Vault raft-snapshot CronJob, CNPG barmanObjectStore backups
(Authentik/n8n/Nextcloud), Nextcloud file-PVC restic sync - all
targeting the new VPS MinIO receiver
See VPS Warm-Standby plan doc for full design rationale.
NetBird router deployments require NET_ADMIN capability, which is
blocked by the default baseline PodSecurity standard. Add
managedNamespaceMetadata with pod-security.kubernetes.io/enforce:
privileged to argocd, frigate, gitlab, grafana, guacamole,
home-assistant, teslamate, and vault ArgoCD Applications.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Network policies in the network-policies/ subdirectory were not being
synced because ArgoCD was not recursing into subdirectories.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The scylla-manager Application was missing the third source for raw
manifests, so the VirtualService in infrastructure/scylla/manifests/
was never deployed. This caused Kiali to flag the ext-authz policy
host as having no matching service entry.
Also removed stale PERMISSIVE PeerAuthentication from authentik
namespace (manually applied, not in git, unnecessary with mesh-wide
STRICT mTLS).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- n8n: Set replicaCount=1 (RWO PVC incompatible with multiple replicas),
add resource limits to satisfy Kyverno policy
- gitlab: Add ignoreDifferences for redis-gitlab-additional service
(port names and selectors managed by Redis operator)
- netbird-operator: Add ExternalSecret for netbird-mgmt-api-key in
netbird namespace and add manifests source to Application
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Gateway API CRDs are already managed by istio; skip them
from the vendored envoy-gateway chart to prevent OutOfSync.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The checksum/lapi-secret annotation on the LAPI Deployment changes
when CrowdSec auto-generates credentials, causing permanent
OutOfSync with ArgoCD ServerSideDiff.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ArgoCD v3.3.3 has a bug where OCI chart revision resolution
doesn't include the chart name in the Docker Hub API path,
causing 401 errors. The chart is already deployed at v1.6.3.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ArgoCD requires the chart field for OCI sources. The repoURL
contains the registry/org path, and chart specifies the chart name.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move chart name into repoURL for OCI registry compatibility.
ArgoCD resolves OCI chart references differently from HTTP repos.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Works with the ignoreDifferences config in argocd-cm to prevent
auto-sync from triggering on controller-managed field changes
(ESO finalizers, HTTPRoute annotations, etc.)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Apps using ServerSideApply without ServerSideDiff causes ArgoCD to
compute diffs client-side, which detects controller-added fields
(finalizers, defaults, annotations) as changes and keeps apps
permanently OutOfSync. ServerSideDiff uses dry-run SSA for accurate
diff computation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The Rook operator reconciles OSD deployments and removes the
expand-bluefs=/bin/true workaround patches needed for BlueStore UUID
mismatch. Keep operator scaled to 0 with manual sync only until the
underlying issue is resolved.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Envoy Gateway moved to OCI registry (docker.io/envoyproxy).
NetBird operator charts moved to https://netbirdio.github.io/helms
with chart name kubernetes-operator.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
