Networker

Re: [Networker] SCSI errors

2003-06-08 08:38:44
Subject: Re: [Networker] SCSI errors
From: "Mark Bradshaw (BTOpenWorld)" <notthehoople AT BTOPENWORLD DOT COM>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Sat, 7 Jun 2003 00:09:53 +0100
Hi Frank,

I've seen a similar problem on a StorageTek L80 with LTO drives and a
Windows machine. In my case the problem was with the settings made in the
NVRAM of the SCSI cards themselves. For whatever reason the cards had been
configured for disk access not tape access and gave sporadic SCSI timeouts
because of this. Might be worth checking....

To check you will need to reboot your system and go into the BIOS of *each*
SCSI card. Don't know anything about the cards you have, but on my Adaptec
cards here you would select:

- Configure/View Host Adapter Settings
--- SCSI Device Configuration

Then check the setting for "Enable Disconnection" for each SCSI ID the card
supports. A setting of "Yes" is a good thing for tapes, a setting of "No"
means that you'll get lots of nasty SCSI timeouts especially if you have
more than one device on each SCSI channel.

Hope this helps....

Mark

> hi again,
>
> It seems that i don't have any time to celebrate any little success,
> since it doesn't take long time for the next problem to appear :-(
>
> Since some weeks i sometimes see SCSI errors in the system log file,
> like these:
>
> Jun  1 12:45:12 r2d2 kernel: scsi : aborting command due to timeout : pid
> 213591 32, scsi2, channel 0, id 0, lun 0 Inquiry 00 00 00 ff 00
>
> In the beginning, there was only scsidev 0,0,0 (the first of five
> LTO ultrium tape devices) that got the error.
> Because i assumed the error in the SCSI card, i disabled this drive
> until i can contact a StorageTek professional to check it.
>
> But now i see SCSI errors on all five tape devices. The tapes are
> distributed to three Tekram DC-390U3W cards.
> So i don't think that the cards are causing the errors, but i don't
> have any clue where to search now.
>
> Additionally to the above entry, i'm getting a lot of these now:
>
> Jun  4 08:40:30 r2d2 kernel: sym53c1010-33-0-<6,*>: target did not report
> SYNC.
> Jun  4 08:40:31 r2d2 kernel: scsi : aborting command due to timeout : pid
> 30973041, scsi5, channel 0, id 4, lun 0 Inquiry 00 00 00 f f 00
> Jun  4 08:40:31 r2d2 kernel: sym53c8xx_abort: pid=30973041
> serial_number=30972996 serial_number_at_timeout=30972996
>
>
> I know this is slightly off-topic, but it's a great problem for me,
> and btw. NetWorker of course refuses to work if a tape devices keeps
> hanging in the 'verifying label' state due to scsi timeouts...
>
> Well, to the hardware:
>
> 3 x Tekram DC-390 U3W Ultra 160 SCSI cards
> 5 x IBM LTO Ultrium drives, builtin to a StorageTek L700 tape library
> 1 x ASUS PR-DLSW Dual Intel Xeon-based motherboard
> 2 x Intel Xeon 2 GHz
> 1 x 1024 MB RAM
> 1 x Redhat Linux 7.3 (Valhalla)
>
> [root@r2d2 root]# cat /proc/scsi/scsi
> Attached devices:
> Host: scsi0 Channel: 00 Id: 05 Lun: 00
> Vendor: easyRAID Model:  F8              Rev: 0001
> Type:   Direct-Access                    ANSI SCSI revision: 03
> Host: scsi0 Channel: 00 Id: 06 Lun: 00
> Vendor: STK      Model: L700             Rev: 0301
> Type:   Medium Changer                   ANSI SCSI revision: 03
> Host: scsi2 Channel: 00 Id: 00 Lun: 00
> Vendor: IBM      Model: ULTRIUM-TD1      Rev: 27Q1
> Type:   Sequential-Access                ANSI SCSI revision: 03
> Host: scsi2 Channel: 00 Id: 01 Lun: 00
> Vendor: IBM      Model: ULTRIUM-TD1      Rev: 27Q1
> Type:   Sequential-Access                ANSI SCSI revision: 03
> Host: scsi2 Channel: 00 Id: 02 Lun: 00
> Vendor: IBM      Model: ULTRIUM-TD1      Rev: 27Q1
> Type:   Sequential-Access                ANSI SCSI revision: 03
> Host: scsi5 Channel: 00 Id: 03 Lun: 00
> Vendor: IBM      Model: ULTRIUM-TD1      Rev: 27Q1
> Type:   Sequential-Access                ANSI SCSI revision: 03
> Host: scsi5 Channel: 00 Id: 04 Lun: 00
> Vendor: IBM      Model: ULTRIUM-TD1      Rev: 27Q1
> Type:   Sequential-Access                ANSI SCSI revision: 03
>
> With kind regards,
>
>       Frank Altpeter
>
> --
> Note: To sign off this list, send a "signoff networker" command via email
> to listserv AT listmail.temple DOT edu or visit the list's Web site at
> http://listmail.temple.edu/archives/networker.html where you can
> also view and post messages to the list.
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

<Prev in Thread] Current Thread [Next in Thread>