RPF-WP-0046: periodic ESO database tokens and a daily renewer

The five dynamic-database ClusterSecretStores use 768h static tokens that
nothing renews; expiry revoked their DB leases on 2026-09-23 and recurs
around 2026-10-25. Kubernetes auth is not a drop-in fix: ESO v0.16.1 revokes
its login token after each reconcile, which revokes the leases it created.

- eso-token-renewer CronJob (ArgoCD draft, no RBAC, mounted Secrets).
- Attended periodic mint script for all five lanes.
- CronJob added to the platform-addons AppProject in git (not yet applied).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 150322@bnt-lap001
Assistant-Session: 16a7b788-374e-4915-a1df-fc87ffd9a5e4

argocd/platform-addons/eso-token-renewer/cronjob.yaml

@@ -0,0 +1,90 @@
+# RPF-WP-0046: daily renew-self for the ESO parent tokens of the five
+# dynamic-database ClusterSecretStores. The Secrets are mounted, not read via
+# the API, so the ServiceAccount has no RBAC and no API token.
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: eso-token-renewer
+  namespace: external-secrets
+  labels:
+    app.kubernetes.io/name: eso-token-renewer
+    app.kubernetes.io/part-of: railiance-platform
+automountServiceAccountToken: false
+---
+apiVersion: batch/v1
+kind: CronJob
+metadata:
+  name: eso-token-renewer
+  namespace: external-secrets
+  labels:
+    app.kubernetes.io/name: eso-token-renewer
+    app.kubernetes.io/part-of: railiance-platform
+spec:
+  schedule: '40 2 * * *'
+  timeZone: Etc/UTC
+  suspend: false
+  concurrencyPolicy: Forbid
+  startingDeadlineSeconds: 3600
+  successfulJobsHistoryLimit: 3
+  failedJobsHistoryLimit: 5
+  jobTemplate:
+    spec:
+      activeDeadlineSeconds: 300
+      backoffLimit: 2
+      ttlSecondsAfterFinished: 604800
+      template:
+        metadata:
+          labels:
+            app.kubernetes.io/name: eso-token-renewer
+        spec:
+          serviceAccountName: eso-token-renewer
+          automountServiceAccountToken: false
+          restartPolicy: Never
+          securityContext:
+            runAsNonRoot: true
+            runAsUser: 65534
+            runAsGroup: 65534
+            fsGroup: 65534
+            seccompProfile:
+              type: RuntimeDefault
+          containers:
+          - name: renew
+            image: python:3.12-slim@sha256:d764629ce0ddd8c71fd371e9901efb324a95789d2315a47db7e4d27e78f1b0e9
+            command: [python3, /worker/renew.py]
+            env:
+            - name: PYTHONDONTWRITEBYTECODE
+              value: '1'
+            - name: BAO_ADDR
+              value: http://openbao.openbao.svc:8200
+            - name: ESO_TOKEN_ROOT
+              value: /var/run/eso-tokens
+            - name: MIN_TTL_SECONDS
+              value: '259200'
+            securityContext:
+              allowPrivilegeEscalation: false
+              readOnlyRootFilesystem: true
+              capabilities:
+                drop: [ALL]
+            resources:
+              requests: {cpu: 10m, memory: 24Mi}
+              limits: {cpu: 100m, memory: 64Mi}
+            volumeMounts:
+            - {name: worker, mountPath: /worker, readOnly: true}
+            - {name: audit-core, mountPath: /var/run/eso-tokens/audit-core, readOnly: true}
+            - {name: canned-prompts, mountPath: /var/run/eso-tokens/canned-prompts, readOnly: true}
+            - {name: core-hub, mountPath: /var/run/eso-tokens/core-hub, readOnly: true}
+            - {name: sbom-nexus, mountPath: /var/run/eso-tokens/sbom-nexus, readOnly: true}
+            - {name: tenant-engine, mountPath: /var/run/eso-tokens/tenant-engine, readOnly: true}
+          volumes:
+          - name: worker
+            configMap: {name: eso-token-renewer-worker, defaultMode: 0444}
+          - name: audit-core
+            secret: {secretName: openbao-audit-core-eso-token, defaultMode: 0440, items: [{key: token, path: token}]}
+          - name: canned-prompts
+            secret: {secretName: openbao-canned-prompts-eso-token, defaultMode: 0440, items: [{key: token, path: token}]}
+          - name: core-hub
+            secret: {secretName: openbao-core-hub-eso-token, defaultMode: 0440, items: [{key: token, path: token}]}
+          - name: sbom-nexus
+            secret: {secretName: openbao-sbom-nexus-eso-token, defaultMode: 0440, items: [{key: token, path: token}]}
+          - name: tenant-engine
+            secret: {secretName: openbao-tenant-engine-eso-token, defaultMode: 0440, items: [{key: token, path: token}]}

argocd/platform-addons/eso-token-renewer/kustomization.yaml

@@ -0,0 +1,16 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+
+# RPF-WP-0046. Read only by the railiance01 Application in
+# argocd/railiance01/drafts/eso-token-renewer.application.yaml. No coulombcore
+# Application points here.
+
+namespace: external-secrets
+
+resources:
+  - cronjob.yaml
+
+configMapGenerator:
+  - name: eso-token-renewer-worker
+    files:
+      - renew.py

argocd/platform-addons/eso-token-renewer/renew.py

@@ -0,0 +1,64 @@
+#!/usr/bin/env python3
+"""Renew the ESO parent tokens that project OpenBao dynamic database leases.
+
+ESO never renews a tokenSecretRef token, and OpenBao revokes every lease a
+token created when that token expires (RPF-WP-0046). Each token Secret is
+mounted read-only at <ESO_TOKEN_ROOT>/<lane>/token. For each lane this calls
+auth/token/renew-self and prints one JSON line with the lane, the resulting
+TTL and an error class. It never prints a token.
+
+Exit 1 if any lane fails or ends below MIN_TTL_SECONDS, so the Job is marked
+failed. A non-periodic token cannot be renewed past its max TTL, so it trips
+the threshold as it nears expiry instead of failing silently.
+"""
+import json
+import os
+from pathlib import Path
+import sys
+import urllib.error
+import urllib.request
+
+ADDR = os.environ.get('BAO_ADDR', 'http://openbao.openbao.svc:8200').rstrip('/')
+ROOT = Path(os.environ.get('ESO_TOKEN_ROOT', '/var/run/eso-tokens'))
+MIN_TTL = int(os.environ.get('MIN_TTL_SECONDS', str(72 * 3600)))
+
+
+def renew(token, opener=urllib.request.urlopen):
+    request = urllib.request.Request(
+        ADDR + '/v1/auth/token/renew-self', data=b'{}', method='POST',
+        headers={'X-Vault-Token': token, 'Content-Type': 'application/json'})
+    with opener(request, timeout=15) as response:
+        auth = json.load(response).get('auth') or {}
+    return {'ttl': auth.get('lease_duration'), 'renewable': auth.get('renewable')}
+
+
+def check_lane(lane_dir, opener=urllib.request.urlopen):
+    result = {'lane': lane_dir.name, 'ok': False}
+    try:
+        token = (lane_dir / 'token').read_text(encoding='utf-8').strip()
+        if not token:
+            return dict(result, error='empty_token')
+        result.update(renew(token, opener))
+    except urllib.error.HTTPError as error:
+        return dict(result, error='http_%s' % error.code)
+    except (urllib.error.URLError, OSError, ValueError) as error:
+        return dict(result, error=type(error).__name__)
+    ttl = result.get('ttl')
+    if not isinstance(ttl, int) or ttl < MIN_TTL:
+        return dict(result, error='ttl_below_threshold')
+    return dict(result, ok=True)
+
+
+def main(opener=urllib.request.urlopen):
+    lanes = sorted(p for p in ROOT.iterdir() if p.is_dir() and not p.name.startswith('.'))
+    if not lanes:
+        print(json.dumps({'ok': False, 'error': 'no_lanes_mounted'}))
+        return 1
+    results = [check_lane(lane, opener) for lane in lanes]
+    for item in results:
+        print(json.dumps(item, sort_keys=True))
+    return 0 if all(item['ok'] for item in results) else 1
+
+
+if __name__ == '__main__':
+    sys.exit(main())