Subject: Re: weird 3.1 SCSI errors
To: None <netbsd-users@netbsd.org>
From: Aaron J. Grier <agrier@poofygoof.com>
List: netbsd-users
Date: 12/06/2007 09:46:55
On Tue, Dec 04, 2007 at 04:35:49PM -0800, John Nemeth wrote:
> I have a 20/40 GB DLT tape drive with a 7 slot changer hooked up
> to my household server running NetBSD 3.1. I'm seeing weird errors and
> timeouts whenever I try to do a backup. The timeouts may be due to the
> machine really being too slow for the tape drive (replacing it is on
> the todo list). Does anybody know what I can do about these errors?
> Would 4.0 fix them?
>
> Dec 3 06:07:41 server /netbsd: aha0: host_stat 13
> Dec 3 06:07:41 server /netbsd: st1(aha0:0:5:0): generic HBA error
> Dec 3 06:08:23 server /netbsd: aha0: host_stat 13
> Dec 3 06:08:23 server /netbsd: st1(aha0:0:5:0): generic HBA error
> Dec 3 06:10:44 server /netbsd: aha0: host_stat 13
> Dec 3 06:10:44 server /netbsd: st1(aha0:0:5:0): generic HBA error
> Dec 3 06:13:20 server /netbsd: st1(aha0:0:5:0): timed out
> Dec 3 06:13:20 server /netbsd: st1(aha0:0:5:0): Check Condition on CDB: 0x1e 00 00 00 00 00
> Dec 3 06:13:20 server /netbsd: SENSE KEY: Aborted Command
> Dec 3 06:13:20 server /netbsd: ASC/ASCQ: Overlapped Commands Attempted
> Dec 3 06:13:20 server /netbsd:
host speed should not be an issue; I have run fast SCSI drives with a
25MHz decstation 5000/200, with appropriate cabling, for instance. I
also ran multiple RAIDframe setups off a 133MHz Pentium, similar to
yours, for a number of years.
> Also, I have 2/4 GB DAT drive in the machine which is giving
> "interesting" errors. Is the drive toast?
>
> Dec 3 09:14:54 server /netbsd: st0(aha0:0:4:0): Check Condition on CDB: 0x0a 00 01 00 00 00
> Dec 3 09:14:54 server /netbsd: SENSE KEY: Hardware Error
> Dec 3 09:14:54 server /netbsd: ASC/ASCQ: ASC 0x44 ASCQ 0xa0
> Dec 3 09:14:54 server /netbsd:
> Dec 3 09:19:11 server /netbsd: st0(aha0:0:4:0): Check Condition on CDB: 0x10 00 00 00 02 00
> Dec 3 09:19:11 server /netbsd: SENSE KEY: Hardware Error
> Dec 3 09:19:11 server /netbsd: INFO FIELD: -737148928
> Dec 3 09:19:11 server /netbsd: ASC/ASCQ: ASC 0x44 ASCQ 0xa0
> Dec 3 09:19:11 server /netbsd:
if your CD and hard drive work fine, that removes problems with the
controller. I'd suspect cabling or termination, the usual SCSI
culprits. if the tape drives are external, are you sure you've disabled
termination on the controller? I think the 1542 has resistor packs to
physically yank. better find a chicken.
--
Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com