Phase 2: deploy Authentik warm standby on the VPS

CNPG operator on the VPS (argocd-apps/vps-standby/cnpg-operator.yaml),
plus a CNPG 'replica cluster' for pg-authentik (infrastructure/
vps-standby/authentik/manifests/cnpg-cluster.yaml) that continuously
replays WAL shipped from home's pg-authentik via VPS MinIO's
cnpg-backups/pg-authentik bucket path - CNPG's native replica-cluster
mode, not a custom restore script like Vault needed (Vault has no
equivalent built-in continuous-replication-into-object-store feature).

The Authentik app itself runs at replicas: 0 in normal operation - the
replica DB is read-only until a deliberate manual promotion
(spec.replica.enabled: false), and a pod trying to write against a
read-only DB would just crashloop uselessly. values.yaml's authentik:
block is copied byte-identical from home's on purpose: the chart
deterministically derives its generated Secret's AUTHENTIK_SECRET_KEY
from these literal values.yaml strings, so both clusters land on the
same key without manually copying it - required since that key decrypts
things stored encrypted in the replicated DB.
This commit is contained in:
Scooby Husky
2026-08-17 22:28:10 -05:00
parent f2213b56b4
commit 1a657ad0a4
5 changed files with 316 additions and 0 deletions
@@ -0,0 +1,68 @@
---
# CNPG "replica cluster" - continuously replays WAL shipped from home's
# pg-authentik cluster (infrastructure/authentik/manifests/cnpg-cluster.yaml)
# via VPS MinIO's cnpg-backups/pg-authentik bucket path. Unlike Vault's
# raft-restore CronJob (Vault has no native continuous-replication-into-
# object-store feature), CNPG's replica-cluster mode is a first-class
# built-in mechanism - no custom restore scripting needed, matches the
# plan's intent directly.
#
# While spec.replica.enabled is true, this cluster is a read-only standby
# in continuous recovery - the -rw service exists but is NOT writable
# (same "nothing accepts writes in normal operation" discipline as Vault/
# Gitea). Promoting it to a real writable primary during an actual
# incident is a deliberate manual step:
# kubectl -n authentik patch cluster pg-authentik --type merge \
# -p '{"spec":{"replica":{"enabled":false}}}'
# then scale the authentik Deployment(s) up from 0.
#
# vps-minio-secret is a plain Secret copied here manually (kubectl, not
# git) from the vps-minio-root-secret in the minio namespace - this
# cluster has no Vault/ESO of its own:
# kubectl -n authentik create secret generic vps-minio-secret \
# --from-literal=accesskey=<vps minio root user> \
# --from-literal=secretkey=<vps minio root password>
apiVersion: postgresql.cnpg.io/v1
kind: Cluster
metadata:
name: pg-authentik
namespace: authentik
spec:
imageName: ghcr.io/cloudnative-pg/postgresql:16
instances: 1
resources:
requests:
memory: "256Mi"
cpu: "25m"
limits:
memory: "1Gi"
cpu: "250m"
storage:
size: 5Gi
storageClass: local-path
bootstrap:
recovery:
source: home-backup
externalClusters:
- name: home-backup
barmanObjectStore:
destinationPath: s3://cnpg-backups/pg-authentik
endpointURL: http://vps-minio.minio.svc.cluster.local:9000
s3Credentials:
accessKeyId:
name: vps-minio-secret
key: accesskey
secretAccessKey:
name: vps-minio-secret
key: secretkey
replica:
enabled: true
source: home-backup
monitoring:
enablePodMonitor: false # no Prometheus on the VPS cluster
@@ -0,0 +1,87 @@
---
# Same idempotent secret-bootstrap pattern as home's
# infrastructure/authentik/manifests/presync-job.yaml, scoped to a
# dedicated ServiceAccount here (no shared argocd-hook-sa exists on the
# vps-standby cluster the way it does at home).
apiVersion: v1
kind: ServiceAccount
metadata:
name: authentik-hook
namespace: authentik
---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
name: authentik-hook
namespace: authentik
rules:
- apiGroups: [""]
resources: ["secrets"]
verbs: ["get", "create"]
- apiGroups: ["postgresql.cnpg.io"]
resources: ["clusters"]
verbs: ["get"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
name: authentik-hook
namespace: authentik
subjects:
- kind: ServiceAccount
name: authentik-hook
namespace: authentik
roleRef:
kind: Role
name: authentik-hook
apiGroup: rbac.authorization.k8s.io
---
apiVersion: batch/v1
kind: Job
metadata:
name: authentik-presync
namespace: authentik
annotations:
argocd.argoproj.io/hook: PreSync
argocd.argoproj.io/hook-delete-policy: BeforeHookCreation
spec:
backoffLimit: 3
template:
spec:
serviceAccountName: authentik-hook
restartPolicy: Never
securityContext:
runAsNonRoot: true
runAsUser: 65534
seccompProfile:
type: RuntimeDefault
containers:
- name: presync
image: alpine/k8s:1.32.13
securityContext:
allowPrivilegeEscalation: false
capabilities:
drop: ["ALL"]
command:
- /bin/bash
- -c
- |
set -euo pipefail
echo "=== Authentik (VPS) PreSync ==="
# NOTE: does NOT wait for pg-authentik to be "Healthy" the way
# home's presync-job does - a replica cluster in continuous
# recovery never reports Healthy in that sense (it's
# perpetually catching up / read-only), and server/worker are
# at replicas: 0 anyway so nothing actually needs the DB yet.
if ! kubectl -n authentik get secret authentik >/dev/null 2>&1; then
echo "Creating authentik secret (literal env:// placeholder -"
echo "matches home's values.yaml exactly, see values.yaml comment)..."
kubectl -n authentik create secret generic authentik \
--from-literal=AUTHENTIK_SECRET_KEY="env://AUTHENTIK_SECRET_KEY"
else
echo "authentik secret already exists"
fi
echo "=== Done ==="