This repository has been archived by the owner on Oct 24, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 519
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
test: add REBOOT_CONTROL_PLANE_NODES E2E config (#3745)
- Loading branch information
1 parent
cb5f0dd
commit 8261389
Showing
4 changed files
with
172 additions
and
43 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
111 changes: 111 additions & 0 deletions
111
test/e2e/kubernetes/workloads/reboot-control-plane-node.yaml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,111 @@ | ||
| # How this works: | ||
| # On every master, a pod will run that will try to get an annotation lock | ||
| # once it gets the lock for its node, it releases it, sleeps a bit and | ||
| # then loops until it can get it again. | ||
| # After it gets it the second time, it will then trigger a node hard reboot | ||
| # (Since force or double force) | ||
| # Then, when the node (master) comes back up, it will again run this pod | ||
| # and it will again see it has the annotation and will release it, letting | ||
| # another node race to get it. | ||
| # Works with ubuntu 16.04 and 18.04 on kubernetes 1.15+ (may work earlier) | ||
| apiVersion: apps/v1 | ||
| kind: DaemonSet | ||
| metadata: | ||
| labels: | ||
| app: reboot-test | ||
| name: reboot-test | ||
| namespace: default | ||
| spec: | ||
| selector: | ||
| matchLabels: | ||
| app: reboot-test | ||
| template: | ||
| metadata: | ||
| labels: | ||
| app: reboot-test | ||
| spec: | ||
| # This restricts this to masters | ||
| affinity: | ||
| nodeAffinity: | ||
| requiredDuringSchedulingIgnoredDuringExecution: | ||
| nodeSelectorTerms: | ||
| - matchExpressions: | ||
| - key: kubernetes.azure.com/role | ||
| operator: In | ||
| values: | ||
| - master | ||
| containers: | ||
| - name: reboot-test | ||
| image: ubuntu:18.04 | ||
| imagePullPolicy: IfNotPresent | ||
| command: | ||
| - bash | ||
| - -c | ||
| - >- | ||
| while [[ $(kubectl --request-timeout 30s annotate namespace ${LOCK_NS} ${LOCK_NAME}=${NODE_ID} 2>&1) != *\(${NODE_ID}\)* ]]; | ||
| do sleep ${CHECK_FREQUENCY}; done; | ||
| kubectl --request-timeout 30s annotate namespace ${LOCK_NS} ${LOCK_NAME}-; | ||
| sleep ${CHECK_DELAY}; | ||
| while [[ $(kubectl --request-timeout 30s annotate namespace ${LOCK_NS} ${LOCK_NAME}=${NODE_ID} 2>&1) != *\(${NODE_ID}\)* ]]; | ||
| do sleep ${CHECK_FREQUENCY}; done; | ||
| echo >/etc/cron.d/reboot-test "* * * * * root /bin/bash -c '/bin/rm -f /etc/cron.d/reboot-test; /sbin/reboot ${FORCE}' >/var/log/RebootTest.log 2>&1"; | ||
| echo "Waiting for the reboot to happen..."; | ||
| sleep 1000 | ||
| env: | ||
| - name: KUBECONFIG | ||
| value: /.kubeconfig | ||
| - name: NODE_ID | ||
| valueFrom: | ||
| fieldRef: | ||
| apiVersion: v1 | ||
| fieldPath: spec.nodeName | ||
| - name: LOCK_NS | ||
| value: default | ||
| - name: LOCK_NAME | ||
| value: RebootTest | ||
| # CHECK_DELAY should be larger than CHECK_FREQUENCY to randomize who reboots | ||
| - name: CHECK_DELAY | ||
| value: 60s | ||
| - name: CHECK_FREQUENCY | ||
| value: 30s | ||
| - name: FORCE | ||
| value: --force | ||
| # For harder reboot, without process kill, use double-force... | ||
| # value: --force --force | ||
| volumeMounts: | ||
| - name: node-crond | ||
| mountPath: /etc/cron.d | ||
| - name: kubectl | ||
| mountPath: /usr/local/bin/kubectl | ||
| readOnly: true | ||
| - name: kubeconfig | ||
| mountPath: /.kubeconfig | ||
| readOnly: true | ||
| - name: certs | ||
| mountPath: /etc/kubernetes/certs | ||
| readOnly: true | ||
| dnsPolicy: ClusterFirst | ||
| nodeSelector: | ||
| beta.kubernetes.io/os: linux | ||
| restartPolicy: Always | ||
| terminationGracePeriodSeconds: 0 | ||
| tolerations: | ||
| - effect: NoSchedule | ||
| operator: Exists | ||
| volumes: | ||
| - hostPath: | ||
| path: /etc/cron.d | ||
| name: node-crond | ||
| - hostPath: | ||
| path: /etc/kubernetes/certs | ||
| name: certs | ||
| - hostPath: | ||
| path: /var/lib/kubelet/kubeconfig | ||
| name: kubeconfig | ||
| - hostPath: | ||
| path: /usr/local/bin/kubectl | ||
| name: kubectl | ||
| updateStrategy: | ||
| rollingUpdate: | ||
| maxUnavailable: 1 | ||
| type: RollingUpdate |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters