Skip to main content
Version: release-0.9

Handle an exception

When an exception occurs during your operation, you can perform the following steps to solve it.


  1. Check the cluster status.

    Take the cluster named mycluster as an example.

    kubectl describe cluster mycluster
  2. Handle the exception according to the status information.

    AbnormalThe cluster can be accessed but exceptions occur in some pods. This might be a mediate status of the operation process and the system recovers automatically without executing any extra operation. Wait until the cluster status changes to Running.
    ConditionsErrorThe cluster is normal but an exception occurs to the condition. It might be caused by configuration loss or exception, which further leads to operation failure. Manual recovery is required.
    FailedThe cluster cannot be accessed. Check the status.message string and get the exception reason. Then manually recover it according to the hints.

    You can check the cluster's status for more information.

Fallback strategies

If the above operations can not solve the problem, try the following steps:

  • Restart this cluster. If the restart fails, you can delete the pod manually.
  • Roll the cluster status back to the status before changes.