RE: RAID5 unusably unstable through 2.6.14

[Salyzyn, Mark]

Martin Drab [mailto:drab@xxxxxxxxxxxxxxxxxxx] sez:
>       sd 0:0:0:0: SCSI error: return code = 0x8000002
>       sda: Current: sense key: Hardware Error
>           Additional sense: Internal target failure
>       Info fld=0x0
>       end_request: I/O error, dev sda, sector <some sector number>

You reported that the Adaptec management software did not indicate the
array was offline in the Adapter, thus these reports come when there is
a unrecoverable (non-redundant) bad block being read from the physical
media. The alternate to this is that such a condition or similar had
existed in the past, and the media bad block was remapped then marked as
inconsistent (as noted, until a write makes it consistent).

However, such conditions do not make the array inaccessible from dd as
you indicate (unless block 0 is the inconsistent block?); thus the array
must have been offline. Too bad we can not reproduce this, the
management applications must have indicated the array was offline (two
drive + failure).

-- Mark
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/