This content has been marked as final. Show 3 replies
Basically there is no easy way to determine a disk failure. In your situation 2 retryable errors on the same block on the same disk should not be of any concearn to you. Format -> Analyze -> Read ... and the issue ir resolved.
Also, there is no threshold on the errors that would indicate a failure. It's more of a common sense in reviewing error counters in iostat considering uptime etc. Also:
If it's unavailable in format - failed
If there is predictive failure analysis > 0 counter in iostat - failed
If it's failed or failing in vxdisk list - failed
If 'need maintenance' in SVM - possibly failed
Other than that, soft/hard/media errors are natural and if you believe that there is too many of them, I suggest contacting support to verify if the disk needs replacement.
Have a look on the docs below:
What is the meaning of the message: <drive not available: formatting> from the Solaris [TM] format(1m) utility? (Doc ID 1005475.1)
How to Interpret Error Counters for the Solaris iostat -E Output (Doc ID 1007250.1)
Solaris[TM] Operating System: High Hard Error value in iostat -E output (Doc ID 1017741.1)
iostat -E : what does "Predictive Failure Analysis" indicate (Doc ID 1003926.1)
SCSI Troubleshooting on Sun Systems (Doc ID 1004694.1)
How to Reset the iostat -E hard/soft/tran Error Counters Without Rebooting (Doc ID 1012731.1)
Hope it helps mate.