Recovering inaccessible or invalid Authentication Keys

Occasionally, an error can occur that requires user intervention. In the event of an error, a cluster fault (referred to as a cluster fault code) will be generated. The two most likely cases are described here.

  1. The cluster is unable to unlock the drives due to a KmipServerFault cluster fault. This can occur when the cluster first boots up and the key server is inaccessible or the required key is unavailable.
    1. Follow the recovery steps in the cluster fault codes (if any).
  2. A sliceServiceUnhealthy fault might be set because the metadata drives have been marked as failed and placed into the "Available" state.
    1. Re-add the drives.
    2. After 3 to 4 minutes, check that the sliceServiceUnhealthy fault has cleared.

See cluster fault code information.

Cluster fault codes