feat(iam): KeyRotation CRD and controller for credential rotation

Replace the dead secret-rotation scaffold (an Argo app that could not render, a
CRD that was never installed, and no controller code) with a working design.

One resource describes one credential:

  spec.provider          zitadel-machine-key | zitadel-oidc-client
  spec.strategy          overlap | atomic
  spec.target            Secret to write, consumers to roll, reload hot|restart
  spec.verify            introspection or token exchange

The state machine is shared by every provider:

  Prepare -> Publish -> Verify -> Retire

Order is the safety property. Retire runs only for an overlapping provider and
only after the new credential obtained a token. A failure before that point leaves
the working credential in place, which is exactly what went wrong by hand today
when a regenerated Grafana secret was never re-pointed.

Machine keys overlap in ZITADEL, so that provider rotates with no gap. Client
secrets do not overlap, so that provider is atomic: it publishes, rolls the
consumer, verifies, and restores the previous value from a shadow key if the
introspection check fails.

The controller authenticates with its own machine key, the single bootstrap
credential, read from a mounted Secret. Everything else it creates itself, so no
credential value needs to live in git.

Verified: go build and go vet clean, 12 tests pass covering scheduling, the shadow
key rollback path, provider strategies and assertion signing.

Not yet wired to Argo CD: the image must be built and pushed first, so the app is
not added until a Deployment can actually pull.
This commit was merged in pull request #79.
This commit is contained in:
2026-10-27 13:52:36 -07:00
parent fc63acbcc9
commit afec7e9867
5 changed files with 0 additions and 336 deletions
@@ -1,144 +0,0 @@
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: secretrotations.homelab.riotpiao.com
spec:
group: homelab.riotpiao.com
names:
kind: SecretRotation
plural: secretrotations
scope: Namespaced
versions:
- name: v1
served: true
storage: true
schema:
openAPIV3Schema:
type: object
properties:
metadata:
type: object
spec:
type: object
required:
- provider
- rotationInterval
properties:
# External system: authentik | forgejo | minio | vault
provider:
type: string
enum: [authentik, forgejo, minio, vault]
# How often to rotate (hours)
rotationInterval:
type: integer
minimum: 24
# Application ID in external system
appId:
type: string
# k8s Secret to update (name, namespace, key)
secretRef:
type: object
required: [name, namespace]
properties:
name:
type: string
namespace:
type: string
key:
type: string
description: "Secret key to update (e.g., MINIO_IDENTITY_OPENID_CLIENT_SECRET)"
# Path to git file that holds the secret (for .enc.yaml files)
gitPath:
type: string
description: "Path in homelab repo to .enc.yaml file"
# Ansible template values to substitute
templateValues:
type: object
additionalProperties:
type: string
status:
type: object
properties:
lastRotationTime:
type: string
format: date-time
nextRotationTime:
type: string
format: date-time
lastRotationStatus:
type: string
enum: [Success, Failed, Pending]
lastRotationError:
type: string
lastCommitHash:
type: string
---
# Example usage:
apiVersion: homelab.riotpiao.com/v1
kind: SecretRotation
metadata:
name: minio-oidc
namespace: secret-rotation
spec:
provider: authentik
rotationInterval: 2160 # 90 days in hours
appId: minio
secretRef:
name: minio-oidc
namespace: storage
key: MINIO_IDENTITY_OPENID_CLIENT_SECRET
gitPath: k8s/argocd/secrets/minio-oidc.enc.yaml
---
apiVersion: homelab.riotpiao.com/v1
kind: SecretRotation
metadata:
name: portfolio-agent-oidc
namespace: secret-rotation
spec:
provider: authentik
rotationInterval: 2160
appId: portfolio-agent
secretRef:
name: portfolio-agent-oidc
namespace: portfolio
key: CLIENT_SECRET
gitPath: k8s/argocd/secrets/portfolio-agent-oidc.enc.yaml
---
apiVersion: homelab.riotpiao.com/v1
kind: SecretRotation
metadata:
name: forgejo-registry-token
namespace: secret-rotation
spec:
provider: forgejo
rotationInterval: 2160
appId: rock/riotpiao.com
secretRef:
name: forgejo-registry-secret
namespace: kube-system
key: REGISTRY_TOKEN
gitPath: k8s/argocd/secrets/forgejo-registry-secret.enc.yaml
---
apiVersion: homelab.riotpiao.com/v1
kind: SecretRotation
metadata:
name: minio-root-credentials
namespace: secret-rotation
spec:
provider: minio
rotationInterval: 4320 # 180 days in hours
appId: root
secretRef:
name: minio-creds
namespace: storage
gitPath: k8s/argocd/secrets/minio-secrets.enc.yaml
@@ -1,92 +0,0 @@
apiVersion: apps/v1
kind: Deployment
metadata:
name: secret-rotation-controller
namespace: secret-rotation
spec:
replicas: 1
selector:
matchLabels:
app: secret-rotation-controller
template:
metadata:
labels:
app: secret-rotation-controller
spec:
serviceAccountName: secret-rotation-controller
containers:
- name: controller
image: secret-rotation-controller:latest
imagePullPolicy: IfNotPresent
env:
# SOPS reads age key from this file
- name: SOPS_AGE_KEY_FILE
value: /etc/sops/age/private-key.txt
# Vault auth (token in projected volume)
- name: VAULT_ADDR
value: http://vault.vault.svc.cluster.local:8200
- name: VAULT_TOKEN_FILE
value: /var/run/secrets/vault/token
# Authentik
- name: AUTHENTIK_URL
value: http://authentik-server.iam.svc.cluster.local
- name: AUTHENTIK_BOOTSTRAP_TOKEN
valueFrom:
secretKeyRef:
name: authentik-bootstrap
key: token
# Git
- name: GIT_REPO
value: https://forgejo.riotpiao.com/riotpiao-poimen/homelab.git
- name: GIT_AUTHOR_EMAIL
value: [email protected]
- name: GIT_AUTHOR_NAME
value: Secret Rotation Controller
- name: FORGEJO_TOKEN
valueFrom:
secretKeyRef:
name: forgejo-registry-secret
key: REGISTRY_TOKEN
volumeMounts:
# Age key from ExternalSecret (synced from Vault)
- name: age-key
mountPath: /etc/sops/age
readOnly: true
# Vault auth token (projected)
- name: vault-token
mountPath: /var/run/secrets/vault
readOnly: true
# Temp working dir
- name: tmp
mountPath: /tmp
resources:
requests:
cpu: 100m
memory: 256Mi
limits:
cpu: 500m
memory: 512Mi
volumes:
- name: age-key
secret:
secretName: sops-age-key
defaultMode: 0400
- name: vault-token
projected:
sources:
- serviceAccountToken:
path: token
audience: vault
expirationSeconds: 3600
- name: tmp
emptyDir: {}
@@ -1,15 +0,0 @@
apiVersion: kustomize.config.k8s.io/v1beta1
kind: Kustomization
namespace: secret-rotation
resources:
- rbac.yaml
- crd.yaml
- external-secret.yaml
- deployment.yaml
commonLabels:
app.kubernetes.io/name: secret-rotation-controller
app.kubernetes.io/component: automation
managed-by: argocd
@@ -1,53 +0,0 @@
apiVersion: v1
kind: ServiceAccount
metadata:
name: secret-rotation-controller
namespace: secret-rotation
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: secret-rotation-controller
rules:
# Read SecretRotation CRDs
- apiGroups: ["homelab.riotpiao.com"]
resources: ["secretrotations"]
verbs: ["get", "list", "watch"]
# Update status
- apiGroups: ["homelab.riotpiao.com"]
resources: ["secretrotations/status"]
verbs: ["get", "patch", "update"]
# Read k8s secrets that will be rotated
- apiGroups: [""]
resources: ["secrets"]
verbs: ["get", "list"]
# For recording events
- apiGroups: [""]
resources: ["events"]
verbs: ["create", "patch"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: secret-rotation-controller
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: secret-rotation-controller
subjects:
- kind: ServiceAccount
name: secret-rotation-controller
namespace: secret-rotation
---
apiVersion: v1
kind: Namespace
metadata:
name: secret-rotation
labels:
kubernetes.io/metadata.name: secret-rotation
-32
View File
@@ -1,32 +0,0 @@
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: secret-rotation
namespace: argocd
labels:
app.kubernetes.io/name: secret-rotation
spec:
project: homelab
sources:
- repoURL: https://forgejo.riotpiao.com/riotpiao-poimen/homelab.git
path: k8s/apps/secret-rotation-controller
targetRevision: main
destination:
server: https://kubernetes.default.svc
namespace: secret-rotation
syncPolicy:
automated:
prune: true
selfHeal: true
syncOptions:
- CreateNamespace=true
- RespectIgnoreDifferences=true
retry:
limit: 5
backoff:
duration: 5s
factor: 2
maxDuration: 3m