Hey guys... from my syslog: May 22 17:05:28 biznatch kernel: scsi : aborting command due to timeout : pid 5591, scsi0, channel 0, id 2, lun 0 Read (10) 00 00 61 01 11 00 00 80 00 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SCSISIGI 0x44, SEQADDR 0x110, SSTAT0 0x0, SSTAT1 0x3 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SG_CACHEPTR 0x0, SSTAT2 0xf, STCNT 0xe8 May 22 17:05:29 biznatch kernel: SCSI host 0 abort (pid 5591) timed out - resetting May 22 17:05:29 biznatch kernel: SCSI bus is being reset for host 0 channel 0. May 22 17:05:30 biznatch kernel: (scsi0:0:2:0) Synchronous at 10.0 Mbyte/sec, offset 15. Can anyone who knows more about SCSI tell me what can cause a command timeout, etc... Do I need a terminator or something? I thought my configuration was right, since I can transfer hundreds of megabytes between errors. Can a problem with terminator configuration even be that subtle? (Forgive a SCSI newbie!) So, the error resolves itself in 7 seconds, and no data is lost, so maybe I don't care. ;) Really, the problem is that it muddies up the console when it happens, and I have to press ^a l. The system is a decade-old experimental dual Pentium-133 running 2.2.21 SMP, with an onboard AIC-7870 (with a BIOS that says "EVALUATION VERSION. DO NOT DISTRIBUTE.") and a 4GB Western Digital Enterprise that someone told me was dead. Heh... okay that sounds pretty bad. So, I guess only pay attention to the terminator question. ;) Thanks. -Chuck
And I thought my 7500 was bad enough! --- Chuck Homic <chuck@vvisions.com> wrote:
Hey guys... from my syslog:
May 22 17:05:28 biznatch kernel: scsi : aborting command due to timeout : pid 5591, scsi0, channel 0, id 2, lun 0 Read (10) 00 00 61 01 11 00 00 80 00 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SCSISIGI 0x44, SEQADDR 0x110, SSTAT0 0x0, SSTAT1 0x3 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SG_CACHEPTR 0x0, SSTAT2 0xf, STCNT 0xe8 May 22 17:05:29 biznatch kernel: SCSI host 0 abort (pid 5591) timed out - resetting May 22 17:05:29 biznatch kernel: SCSI bus is being reset for host 0 channel 0. May 22 17:05:30 biznatch kernel: (scsi0:0:2:0) Synchronous at 10.0 Mbyte/sec, offset 15.
Can anyone who knows more about SCSI tell me what can cause a command timeout, etc... Do I need a terminator or something? I thought my configuration was right, since I can transfer hundreds of megabytes between errors. Can a problem with terminator configuration even be that subtle? (Forgive a SCSI newbie!)
So, the error resolves itself in 7 seconds, and no data is lost, so maybe I don't care. ;) Really, the problem is that it muddies up the console when it happens, and I have to press ^a l.
The system is a decade-old experimental dual Pentium-133 running 2.2.21 SMP, with an onboard AIC-7870 (with a BIOS that says "EVALUATION VERSION. DO NOT DISTRIBUTE.") and a 4GB Western Digital Enterprise that someone told me was dead. Heh... okay that sounds pretty bad.
So, I guess only pay attention to the terminator question. ;)
Thanks.
-Chuck
_______________________________________________ Wlug mailing list Wlug@mail.wlug.org http://mail.wlug.org/mailman/listinfo/wlug
===== - __________________________________________________ Do You Yahoo!? LAUNCH - Your Yahoo! Music Experience http://launch.yahoo.com
Well considering the hardware, I'm impressed that you've got it cooking at all. As far as termination goes, there's really only one place to check. It sounds to me like you've got a single scsi device on that chain. Since the AIC 78xx auto terminate their end by default, that shouldn't be a problem, so all you have to do is double check that the drive itself is either properly set up to terminate the chain or that there is a terminator after it. You can also play with the powered termination if that drive supports it, but I honestly have never been able to figure out when that is necesary, and when it isn't. Just as an FYI ... I had the exact same problem with my 386 back in the day, where it would just freak out every so often. After moving that drive off of the wimpy little trantor card that it was on, and putting it in a real case (with better cooling) I never recieved the errors again. So to sum up. Check to make sure the drive is staying cool. Check to make sure the drive is terminated. Play with termination power to see if it makes that problem go away. And when all else fails, say f**k it, and just live with the random log messages. Hope that helps a little. Cheers, Lee On Wed, 22 May 2002, Chuck Homic wrote:
Hey guys... from my syslog:
May 22 17:05:28 biznatch kernel: scsi : aborting command due to timeout : pid 5591, scsi0, channel 0, id 2, lun 0 Read (10) 00 00 61 01 11 00 00 80 00 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SCSISIGI 0x44, SEQADDR 0x110, SSTAT0 0x0, SSTAT1 0x3 May 22 17:05:28 biznatch kernel: (scsi0:0:2:0) SG_CACHEPTR 0x0, SSTAT2 0xf, STCNT 0xe8 May 22 17:05:29 biznatch kernel: SCSI host 0 abort (pid 5591) timed out - resetting May 22 17:05:29 biznatch kernel: SCSI bus is being reset for host 0 channel 0. May 22 17:05:30 biznatch kernel: (scsi0:0:2:0) Synchronous at 10.0 Mbyte/sec, offset 15.
Can anyone who knows more about SCSI tell me what can cause a command timeout, etc... Do I need a terminator or something? I thought my configuration was right, since I can transfer hundreds of megabytes between errors. Can a problem with terminator configuration even be that subtle? (Forgive a SCSI newbie!)
So, the error resolves itself in 7 seconds, and no data is lost, so maybe I don't care. ;) Really, the problem is that it muddies up the console when it happens, and I have to press ^a l.
The system is a decade-old experimental dual Pentium-133 running 2.2.21 SMP, with an onboard AIC-7870 (with a BIOS that says "EVALUATION VERSION. DO NOT DISTRIBUTE.") and a 4GB Western Digital Enterprise that someone told me was dead. Heh... okay that sounds pretty bad.
So, I guess only pay attention to the terminator question. ;)
Thanks.
-Chuck
_______________________________________________ Wlug mailing list Wlug@mail.wlug.org http://mail.wlug.org/mailman/listinfo/wlug
-- Lee Keyser-Allen lkeyser@wpi.edu WPI CS 2002
From: Chuck Homic <chuck@vvisions.com>
Hey guys... from my syslog:
May 22 17:05:28 biznatch kernel: scsi : aborting command due to timeout <etc> Can anyone who knows more about SCSI tell me what can cause a command timeout, etc...
That looks a lot like the messages I got from a drive that was about to die. In my case, timeouts got more and more frequent, then the drive started making rattling-grinding sounds and then quit working at all. At the time I was not happy to lose the drive, but in retrospect I appreciate the warning. Because it took a few days to totally die, I was able to copy the essential data from it to a new place. I strongly suggest that you not turn on this machine except to do back ups. Save your files!
Do I need a terminator or something?
Could be. Bad termination will make a SCSI device act flakey.
So, the error resolves itself in 7 seconds, and no data is lost,
...yet...
and a 4GB Western Digital Enterprise that someone told me was dead. Heh... okay that sounds pretty bad.
Reports of its death may have been premature, but don't count on a long reprieve. After it is backed up you can play with terminators and configurations. -- -- Keith Wright <kwright@free-comp-shop.com> Programmer in Chief, Free Computer Shop <http://www.free-comp-shop.com> --- Food, Shelter, Source code. ---
Hey guys, I'm about to acquire an Intel 1U PIII-650 256MB 10gig server, used (racked & left) for $250, (*cheer*), and I had intended to use it for my audio recording purposes. My first idea was ProTools and Win2K on it, but as I'd always love a reason to go full Linux, I wondered if there was a quality, (hopefully free), recording/audio editing software option out there. Suggestions would be appreciated! :) -Sands _______________________________________________ "daddy, what does 'formatting hard disk' mean?" __________________________________________________ Do You Yahoo!? LAUNCH - Your Yahoo! Music Experience http://launch.yahoo.com
On Thu, May 23, 2002 at 02:33:09AM -0400, Keith Wright wrote:
That looks a lot like the messages I got from a drive that was about to die. In my case, timeouts got more and more frequent, then the drive started making rattling-grinding sounds
Hm... now that you mention it... grinding...
Save your files!
No problem, I'm just using it to cache a bunch of MP3 CD's, so I don't have to swap discs, so I can let it burn.
and a 4GB Western Digital Enterprise that someone told me was dead. Heh... okay that sounds pretty bad.
Reports of its death may have been premature, but don't count on a long reprieve. After it is backed up you can play with terminators and configurations.
I figure the level of failure it's reached so far is enough to confound a Windows machine. (I don't know, does Windows throw around SCSI bus resets when things go awry?) But I'll have to dig up my bucket o' SCSI crap and see if I can't try out a few terminators, or look up the drive on the web site and see if it's got any termination jumpers or whatnot. Thanks. -Chuck
On Sun, May 26, 2002 at 01:09:56PM -0400, Chuck Homic wrote:
On Thu, May 23, 2002 at 02:33:09AM -0400, Keith Wright wrote:
That looks a lot like the messages I got from a drive that was about to die. In my case, timeouts got more and more frequent, then the drive started making rattling-grinding sounds
Hm... now that you mention it... grinding...
now that i think about it, i had the same probablem. in order to fix it i turned the speed of the scsi bus down. You can give that a try, by going into the SCSI-Select utility (ctl-a on boot), and turn down the scsi bus speed. --brad
participants (7)
-
Brad Noyes
-
Chuck Homic
-
Jeremy
-
Josh Huber
-
Keith Wright
-
Lee Keyser-Allen
-
Sands Fish