Sounds like the Kerberos ticket expired in the local cache.
Can you determine later in the log file if that WAL deletion was retried and succeeded/failed again? Is this something that you repeatedly see? Does the tserver keep running?
On 3/16/14, 6:38 PM, Kepner, Jeremy - 0553 - MITLL wrote:
Anyone ever seen these errors:
Error attempting to delete write-ahead log
e1c09270-b817-4fac-804d-f2e70c9690ac: org.apache.hadoop.ipc.RemoteException:
GSS initiate failed
