Troubleshooting
Cluster Sleep/Wake
When ordering this demo from demo.redhat.com, you cluster will be woken up from a stopped state. Once startup is complete, you should wait around a further 15 minutes for the cluster to fully initialize. There are occasions when certain components in the cluster are not initialized in a healthy state and the following issues occur:
Webhook not triggering pipeline run
You will notice that any source code changes, tagging and releasing will not trigger a pipeline. To confirm this issue, navigate to your software components xxx-dev project and go to Workloads → Pods then locate the event listener pod with the name el-<component>-el*. If you select the Logs tab, you should notice a certificate error while communicating tekton-trigger-core-interceptors service. To fix this issue, switch the project to openshift-pipelines and go to Workloads → Pods. Find the pod tekton-triggers-core-interceptors-* and delete it. This action should recreate the pod and should resolve this issue. Attempt to trigger the pipeline again.
If there are no logs in the event listener indicating a certificate issue, this may be caused by a timing issue i.e. you may have triggered your pipeline before the webhook was successfully created. To ensure that the webhook was created, open your software component on Developer Hub and ensure you are on the Overview tab. On the bottom right, under Deployment Summary confirm that the ArgoCD App <component-name>-dev-build has a Sync status of Synced and Health status of Healthy before triggering your pipeline.