Troubleshooting

Most Klutch components run in the Control Plane Cluster, where detailed logs are available. Developers can review konnector logs and inspect any errors synchronized between the control plane and the application cluster, as well as the logs and events of their deployed resources.

Symptom: Resource never reaches Ready

1. Inspect the resource:

   kubectl describe <resource-name> <name> -n <ns>

   Look for warning events or failed conditions.

2. Check the full status:

   kubectl get <resource-name> <name> -n <ns> -o yaml

   Review the status.conditions and any references to the backing composite resource.

3. Check the konnector logs:

   kubectl logs -l <konnector-label> -n <konnector-namespace>

   Look for authentication or connectivity issues between your cluster and the control plane.

Symptom: instance is not Ready after configuring config.parameterGroup

1. Inspect the resource:

   kubectl describe <resource-name> <name> -n <ns>
   kubectl get <resource-name> <name> -n <ns> -o yaml

   Look for warning events or failed conditions.

2. Verify parameter group values and units:

   Ensure values match supported types and ranges (for example MiB fields and integer-only fields).

3. Re-apply with corrected values:

   kubectl apply -f <instance-manifest>.yaml

For supported parameter fields and examples, see Configure Parameter Groups.