ATA error messages

[Lowell Gilbert]

Martin Schweizer <pcservi at spectraweb.ch> writes:

> Since two days I log in /var/log/messages the following errors. What does they 
> mean (I now, somthing with my first harddrive... :-) )?

Yep.

> My System: 
> FreeBSD saturn.pcs.ms 4.8-RELEASE FreeBSD 4.8-RELEASE #0: Thu May 15 08:23:22 CEST 
> 2003     martin at saturn.spectraweb.ch:/usr/obj/usr/src/sys/KERNEL48  i386
> 
> /var/log/messages:
> Oct 27 09:36:59 saturn /kernel: ata0: resetting devices .. done
> Oct 27 09:36:59 saturn /kernel: ad0: READ command timeout tag=0 serv=0 - resetting
> Oct 27 09:36:59 saturn /kernel: ata0: resetting devices .. done
> Oct 27 09:44:05 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97360, size: 57344
> Oct 27 09:44:47 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 95680, size: 32768
> Oct 27 09:44:50 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97880, size: 49152
> Oct 27 09:44:54 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97360, size: 57344
> Oct 27 09:44:59 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 95680, size: 32768
> Oct 27 09:44:59 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97880, size: 49152
> Oct 27 09:45:02 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97360, size: 57344
> Oct 27 09:45:03 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 95680, size: 32768
> Oct 27 09:45:03 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 97880, size: 49152
> Oct 27 09:46:30 saturn /kernel: swap_pager: indefinite wait buffer: device: #ad/0x30001, blkno: 51216, size: 4096
> Oct 28 09:35:45 saturn /kernel: ata0: resetting devices .. done
> Oct 28 09:35:49 saturn /kernel: ad0: READ command timeout tag=0 serv=0 - resetting
> Oct 28 09:35:49 saturn /kernel: ata0: resetting devices .. done

These are all various kinds of reports of the disk not reporting back
by the time the system expected it to do so.  The usual first culprit
to check is the drive cable.  I've been getting the command timeouts
on -STABLE for the last few weeks, and I'm figuring that the drive is
probably on its way out -- but that it might take months to get there.