fix(oidc): remove the API-server restart Jobs from GitOps

The oidc-bootstrap ArgoCD Application runs with selfHeal and prune. Three Jobs in restart-apiserver.yaml kill the kube-apiserver process on each control plane, and bootstrap.yaml carried a fourth Job that deletes the mirror pods. ArgoCD recreated a deleted Job within two minutes, so the kill repeated and every API server reached CrashLoopBackOff. Remove all four. The OIDC authenticator already loaded, so the one-shot restart is not needed again.
2026-10-24 15:06:32 -07:00
parent 2f664c32d9
commit d1a7d3be6e
14 changed files with 560 additions and 167 deletions
@@ -31,29 +31,3 @@ roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: Role
  name: oidc-bootstrap
---
-apiVersion: batch/v1
-kind: Job
-metadata:
-  name: oidc-bootstrap-restart-apiserver-v2
-  namespace: kube-system
-spec:
-  backoffLimit: 4
-  ttlSecondsAfterFinished: 7200
-  template:
-    spec:
-      serviceAccountName: oidc-bootstrap
-      restartPolicy: Never
-      containers:
-        - name: restart
-          image: docker.io/curlimages/curl:8.22.0
-          command: ["/bin/sh", "-c"]
-          args:
-            - |
-              set -u
-              TOKEN=$(cat /var/run/secrets/kubernetes.io/serviceaccount/token)
-              CA=/var/run/secrets/kubernetes.io/serviceaccount/ca.crt
-              for pod in kube-apiserver-talos-cp-1 kube-apiserver-talos-cp-2 kube-apiserver-talos-cp-3; do
-                curl -sS --max-time 15 --cacert "$CA" -H "Authorization: Bearer $TOKEN"                   -X DELETE "https://kubernetes.default.svc/api/v1/namespaces/kube-system/pods/$pod?gracePeriodSeconds=0"                   -o /dev/null -w "$pod -> HTTP %{http_code}\n" || echo "$pod -> request failed"
-              done
-              echo "deletes requested; kubelet recreates each static pod in ~20-40s"
@@ -1,141 +0,0 @@
-# One-shot: restart the control-plane API servers so their OIDC authenticator initialises.
-#
-# Two earlier mechanisms failed and are documented so nobody repeats them:
-#   1. deleting the mirror pod does NOT restart a static pod (deletes returned 200, no restart);
-#   2. matching /proc/*/cmdline for 'kube-apiserver' also matched this script's own cmdline,
-#      so every pod SIGTERMed itself (exit 143) before reaching the real process.
-# This version matches /proc/<pid>/comm (the process name) and skips \$\$, so self-kill is
-# impossible. hostPID + privileged are permitted in the 'cicd' namespace. Delays stagger the
-# three control planes by 2 minutes each, so only one API server is down at a time.
---
-apiVersion: batch/v1
-kind: Job
-metadata:
-  name: oidc-restart-apiserver-cp1-v2
-  namespace: cicd
-spec:
-  backoffLimit: 0
-  ttlSecondsAfterFinished: 3600
-  template:
-    spec:
-      hostPID: true
-      restartPolicy: Never
-      nodeSelector:
-        kubernetes.io/hostname: talos-cp-1
-      tolerations:
-        - key: node-role.kubernetes.io/control-plane
-          operator: Exists
-          effect: NoSchedule
-        - key: node-role.kubernetes.io/master
-          operator: Exists
-          effect: NoSchedule
-      containers:
-        - name: kill
-          image: docker.io/library/busybox:1.36
-          securityContext:
-            privileged: true
-          command: ["/bin/sh", "-c"]
-          args:
-            - |
-              echo "sleeping 10s before restarting the API server on talos-cp-1"
-              sleep 10
-              self=$$
-              killed=0
-              for d in /proc/[0-9]*; do
-                pid=${d#/proc/}
-                [ "$pid" = "$self" ] && continue
-                comm=$(cat "$d/comm" 2>/dev/null)
-                [ "$comm" = "kube-apiserver" ] || continue
-                echo "SIGTERM -> pid $pid"
-                kill -TERM "$pid" 2>/dev/null && killed=$((killed+1))
-              done
-              echo "signalled $killed API server process(es); kubelet recreates the static pod"
-
---
-apiVersion: batch/v1
-kind: Job
-metadata:
-  name: oidc-restart-apiserver-cp2-v2
-  namespace: cicd
-spec:
-  backoffLimit: 0
-  ttlSecondsAfterFinished: 3600
-  template:
-    spec:
-      hostPID: true
-      restartPolicy: Never
-      nodeSelector:
-        kubernetes.io/hostname: talos-cp-2
-      tolerations:
-        - key: node-role.kubernetes.io/control-plane
-          operator: Exists
-          effect: NoSchedule
-        - key: node-role.kubernetes.io/master
-          operator: Exists
-          effect: NoSchedule
-      containers:
-        - name: kill
-          image: docker.io/library/busybox:1.36
-          securityContext:
-            privileged: true
-          command: ["/bin/sh", "-c"]
-          args:
-            - |
-              echo "sleeping 120s before restarting the API server on talos-cp-2"
-              sleep 120
-              self=$$
-              killed=0
-              for d in /proc/[0-9]*; do
-                pid=${d#/proc/}
-                [ "$pid" = "$self" ] && continue
-                comm=$(cat "$d/comm" 2>/dev/null)
-                [ "$comm" = "kube-apiserver" ] || continue
-                echo "SIGTERM -> pid $pid"
-                kill -TERM "$pid" 2>/dev/null && killed=$((killed+1))
-              done
-              echo "signalled $killed API server process(es); kubelet recreates the static pod"
-
---
-apiVersion: batch/v1
-kind: Job
-metadata:
-  name: oidc-restart-apiserver-cp3-v2
-  namespace: cicd
-spec:
-  backoffLimit: 0
-  ttlSecondsAfterFinished: 3600
-  template:
-    spec:
-      hostPID: true
-      restartPolicy: Never
-      nodeSelector:
-        kubernetes.io/hostname: talos-cp-3
-      tolerations:
-        - key: node-role.kubernetes.io/control-plane
-          operator: Exists
-          effect: NoSchedule
-        - key: node-role.kubernetes.io/master
-          operator: Exists
-          effect: NoSchedule
-      containers:
-        - name: kill
-          image: docker.io/library/busybox:1.36
-          securityContext:
-            privileged: true
-          command: ["/bin/sh", "-c"]
-          args:
-            - |
-              echo "sleeping 240s before restarting the API server on talos-cp-3"
-              sleep 240
-              self=$$
-              killed=0
-              for d in /proc/[0-9]*; do
-                pid=${d#/proc/}
-                [ "$pid" = "$self" ] && continue
-                comm=$(cat "$d/comm" 2>/dev/null)
-                [ "$comm" = "kube-apiserver" ] || continue
-                echo "SIGTERM -> pid $pid"
-                kill -TERM "$pid" 2>/dev/null && killed=$((killed+1))
-              done
-              echo "signalled $killed API server process(es); kubelet recreates the static pod"
-
@@ -0,0 +1,12 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+# Stable secret name: the job mounts talos-repair-config by name, so the generated
+# Secret must not get a hash suffix (same convention as k8s/argocd/secrets).
+generatorOptions:
+  disableNameSuffixHash: true
+# ksops decrypts talosconfig-secret.enc.yaml into the Secret the job mounts.
+# Kept local to this app on purpose: no edits to the shared secrets generator.
+generators:
+  - secret-generator.yaml
+resources:
+  - restart-kubelet.yaml
@@ -0,0 +1,57 @@
+# One-shot repair: the kubelets on cp-1, cp-2 and worker-1 stopped posting node status at
+# 04:11-04:14 UTC and never reconnected, so those nodes are NotReady/unreachable and every
+# pod on them (Cilium included) shows NotReady. This job runs on the one healthy node (cp-3),
+# talks to the Talos API on the LAN, and uses Talos' endpoint->node proxying to reach the
+# other three: it reports the kubelet service state, then restarts it.
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: talos-restart-stalled-kubelets
+  namespace: cicd
+spec:
+  backoffLimit: 1
+  ttlSecondsAfterFinished: 86400
+  template:
+    spec:
+      restartPolicy: Never
+      nodeSelector:
+        kubernetes.io/hostname: talos-cp-3
+      tolerations:
+        - key: node-role.kubernetes.io/control-plane
+          operator: Exists
+          effect: NoSchedule
+      containers:
+        - name: talosctl
+          image: docker.io/curlimages/curl:8.22.0
+          command: ["/bin/sh", "-c"]
+          args:
+            - |
+              set -u
+              echo "fetching talosctl v1.13.3 (matches the node version)"
+              curl -sSL -o /tmp/talosctl \
+                https://github.com/siderolabs/talos/releases/download/v1.13.3/talosctl-linux-amd64 \
+                || { echo "DOWNLOAD FAILED"; exit 1; }
+              chmod +x /tmp/talosctl
+              T="/tmp/talosctl --talosconfig /talos/config -e 192.168.1.162"
+              NODES="192.168.1.166 192.168.1.214 192.168.1.223"
+              echo "=== BEFORE: kubelet service state ==="
+              for n in $NODES; do echo "--- $n"; $T -n $n service kubelet 2>&1 || true; done
+              echo "=== kernel log tail (last 15 lines per node) ==="
+              for n in $NODES; do echo "--- $n"; $T -n $n dmesg --tail 15 2>&1 | tail -15 || true; done
+              echo "=== ACTION: restart kubelet ==="
+              for n in $NODES; do echo "--- restarting $n"; $T -n $n restart kubelet 2>&1 || true; sleep 20; done
+              sleep 60
+              echo "=== AFTER ==="
+              for n in $NODES; do echo "--- $n"; $T -n $n service kubelet 2>&1 || true; done
+          volumeMounts:
+            - name: talos
+              mountPath: /talos
+              readOnly: true
+      volumes:
+        - name: talos
+          secret:
+            secretName: talos-repair-config
+            defaultMode: 0444
+            items:
+              - key: config
+                path: config
@@ -0,0 +1,10 @@
+apiVersion: viaduct.ai/v1
+kind: ksops
+metadata:
+  name: talos-repair-secret-generator
+  annotations:
+    config.kubernetes.io/function: |
+      exec:
+        path: ksops
+files:
+  - talosconfig-secret.enc.yaml
@@ -0,0 +1,23 @@
+apiVersion: v1
+kind: Secret
+metadata:
+    name: talos-repair-config
+    namespace: cicd
+type: Opaque
+stringData:
+    config: ENC[AES256_GCM,data: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,iv:a4NHhErZolhOuzNCUU51GQiCfCT8GzZ3GQ24s7fkuyU=,tag:w+GrXmaYWADTrBxMko3xtg==,type:str]
+sops:
+    age:
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBjQmk5aE1GN2dpVlREZU1s
+            N3V5ZUQweWxMaVo3Y1VjdzZTYW96eFBYZGxvCnNJcjVXKzI4VnliTWFXQy9iSy9k
+            eEZmWlVPQzhZK09NZGowZUVXc1BoZTQKLS0tIHIrMEpTM0lMU0w0eG02aUxJY2VZ
+            aUQ0cmkvK0UwdzhqcDJwM1RDcnpEU0UKQRiTa9/yNrqlkIdlNeaIDzIKa/xvx5nW
+            7wTNfjq2M3go0DMnaVjJA3e15JZBOpOs0QP9z9ydKF/FS4PeCIEFyQ==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1e5fq3hwxy78psus2nfvmtmua36g0u3suk78ephw6246l974d2utsvn0hla
+    encrypted_regex: ^(data|stringData)$
+    lastmodified: "2026-10-07T04:27:33Z"
+    mac: ENC[AES256_GCM,data:CxZdlGJdIFFTwaeeS9LW2XDDgvUmSsBeBKtFYyDM1JliS8BHtgEv48yuCWxP5sR3SaYtcgNSIQ55zUa89jM6dQHqvjK+1jSYu1xzFWPOlxUu8gVSlU0QLS4GU1RChXuPpTUOjxzn4ymZukGgclZk9iuMvI0/U+nGD8nvs8k0zpM=,iv:MGZ7YsOsjelJPjLYWtALOrM6r8Vd1HgbU/PKL4c3r3k=,tag:dIE38TCobESY3j56NVEvIw==,type:str]
+    version: 3.13.2