Identifying Bad Drives
35
•
Long Read or Write Command Times - There is new functionality with
v 3.5 and later that rewrites a block whenever there is a long read or write.
This has been shown to resolve the issue in many cases. Therefore, the
occasional long read or long write can be safely ignored. However, if you
see a disk that is reporting more that three long command times in a week,
it should be replaced.
n
It is possible that once a long I/O happens, other requests to that drive are
also held off with a long I/O message. Make sure that the three long I/Os over
the week are separate long I/Os, and not just multiple reports of the same long
I/O from the same drive.
•
STATUS_IO_DEVICE_ERROR - This is a serious hardware error and
the disk should be replaced.
n
If you take the disk offline from the controller manually, you might see the
STATUS_IO_DEVICE_ERROR error.
