Troubleshooting
You may receive errors while installing or uninstalling on Kubernetes.
Resolve installation errors
If you receive errors (like the following example) during the installation of the catalog solution, these errors might be caused by a previous installation of the catalog solution that was not cleanly uninstalled (or failed to uninstall):
Error: rendered manifests contain a resource that already exists.
Unable to continue with install: existing resource conflict: ...
...
...
Perform the following step to delete existing resources manually before re-installing:
- Enter the following command, replacing
<ldc>with your Helm instance name:kubectl delete all -l "app.kubernetes.io/instance=<ldc>-test" -l "release=<ldc>"
Resolve uninstallation errors
If using the
helm uninstall <instance-name> uninstall solution does not work as expected, you can uninstall without hooks, then manually delete orphaned or incomplete jobs manually.Perform the following steps to uninstall without hooks, replacing <ldc> with your Helm instance name:
Procedure
Use the following example command to uninstall:
helm uninstall <ldc> --no-hooksUse the following example command to determine if any orphaned or incomplete jobs exist:
kubectl get jobs -o name | grep <ldc>Use the
kubectl deletecommand to delete any orphaned or incomplete jobs.