Scsiformat does remapping of bad blocks. You can see the bad blocks in the disks Grown Defect List afterward (if you have a tool to read the GDL). Your disk is good, some bad blocks can be expected during the useful life of the disk. The key is to evaluate when "some" is too many (rate of increase), but 4 isn't too many. Andy -----Original Message----- From: Louis-David Mitterrand [mailto:vindex@apartia.org] Sent: Thursday, October 10, 2002 6:29 AM To: linux-raid@vger.kernel.org Subject: scsiformat and badblocks (was: Re: repeatable crash during raid5 rebuild on 2.4.19-smp) On Sun, Oct 06, 2002 at 12:16:37PM +0200, Louis-David Mitterrand wrote: > - "mdadm /dev/md1 --add /dev/sda2" starts OK, says 15 minutes to go, > when at 25% rebuild suddenly it stops with "md_do_sync() caught > signal, exiting" Following up to myself. Could bad blocks on one of the raid5 partition disk have caused this problem? I just found out that /dev/sdb has 4 four badblocks, the following messages appearing in the syslog during the "badblocks /dev/sdb" command: Oct 10 08:51:12 uruk kernel: scsi0: ERROR on channel 0, id 1, lun 0, CDB: Read ( 10) 00 01 20 ba 30 00 00 c4 00 Oct 10 08:51:12 uruk kernel: Info fld=0x120ba48, Current sd08:10: sense key Medium Error Oct 10 08:51:12 uruk kernel: Additional sense indicates Read retries exhausted Oct 10 08:51:12 uruk kernel: I/O error: dev 08:10, sector 18922056 So I tried a "scsiformat /dev/sdb" and now the disks seems to have no more bad blocks. Can a low-level format fix badblocks on a scsi disk (Fujitsu MAM3376)? Should I trust that disk or send it back? Cheers, - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
