scsi disk problem or kernel wierdness

scsi disk problem or kernel wierdness

Post by Jim Smal » Mon, 16 Dec 1996 04:00:00



I have a adaptec 1542cf, an internal 1 gig drive
an externel space heater (er uh, seagate baracuda 2gig), and a
7 cdrom changer (ncec 4x it's at the end of the chain with termination
turned on)

I keep getting these

Dec 15 21:44:55 ex500 kernel: scsi0: MEDIUM ERROR on channel 0, id 0,
lun 0, CDB: 0x0a 0e 92 c9 08 00
Dec 15 21:44:55 ex500 kernel: Current error sd08:05: sns = f0  3
Dec 15 21:44:55 ex500 kernel: ASC=10 ASCQ= 0
Dec 15 21:44:55 ex500 kernel: Raw sense data:0xf0 0x00 0x03 0x00 0x0e
0x92 0xce 0x28 0x00 0x00 0x00 0x00 0x10 0x00 0x00 0x80
Dec 15 21:44:55 ex500 kernel: scsidisk I/O error: dev 08:05, sector
443084
Dec 15 23:23:00 ex500 kernel: scsi0: MEDIUM ERROR on channel 0, id 0,
lun 0, CDB: 0x0a 0e 92 c9 08 00
Dec 15 23:23:00 ex500 kernel: Current error sd08:05: sns = f0  3
Dec 15 23:23:00 ex500 kernel: ASC=10 ASCQ= 0
Dec 15 23:23:00 ex500 kernel: Raw sense data:0xf0 0x00 0x03 0x00 0x0e
0x92 0xce 0x28 0x00 0x00 0x00 0x00 0x10 0x00 0x00 0x80
Dec 15 23:23:00 ex500 kernel: scsidisk I/O error: dev 08:05, sector
443084

Dec 15 20:08:41 ex500 kernel: scsidisk I/O error: dev 08:05, sector
120412
Dec 15 20:40:41 ex500 kernel: scsidisk I/O error: dev 08:05, sector
216982
Dec 15 20:40:42 ex500 kernel: scsidisk I/O error: dev 08:05, sector
216984
Dec 15 20:41:21 ex500 kernel: scsidisk I/O error: dev 08:05, sector
216984
Dec 15 20:45:16 ex500 kernel: scsidisk I/O error: dev 08:05, sector
218396
Dec 15 21:04:44 ex500 kernel: scsidisk I/O error: dev 08:05, sector
443084
Dec 15 21:44:55 ex500 kernel: scsidisk I/O error: dev 08:05, sector
443084
Dec 15 23:23:00 ex500 kernel: scsidisk I/O error: dev 08:05, sector
443084

unforunately, I do not know if this started when I bought my cdchanger,
or when I switched over to 2.0.x (I was previously using 1.2.13)

The errors are serious enough that when I'm ftp-ing large files via my
ppp link,  they have ERRORS in them (won't gunzip, crc errors)

and I suspect the same thing is happening when I'm making kernels.
basically, either theres

1) something wrong with the 2.0.x kernel and adaptec 1542 driver

2) my internal hard drive is diying (3 years old)

3) I have a termination problem

4) I have a cabling problem

---
[this gets even more annoying as I'm currently trying to debug why
a new machine I just loaded with slackware's distribution keeps on
giving me signal 11's when I try to compile the kernel (as well as
when I run java :-(]

please e-mail a reply

--
nethole#1   --- Resident Intelligent Donkey ---
http://www.nethole.com
WAH #1

 
 
 

scsi disk problem or kernel wierdness

Post by Mark Sutto » Thu, 19 Dec 1996 04:00:00



> I have a adaptec 1542cf, an internal 1 gig drive
> an externel space heater (er uh, seagate baracuda 2gig), and a
> 7 cdrom changer (ncec 4x it's at the end of the chain with termination
> turned on)

> I keep getting these

[actual error messages snipped]
Quote:

> unforunately, I do not know if this started when I bought my cdchanger,
> or when I switched over to 2.0.x (I was previously using 1.2.13)

> The errors are serious enough that when I'm ftp-ing large files via my
> ppp link,  they have ERRORS in them (won't gunzip, crc errors)

> and I suspect the same thing is happening when I'm making kernels.
> basically, either theres

> 1) something wrong with the 2.0.x kernel and adaptec 1542 driver

> 2) my internal hard drive is diying (3 years old)

> 3) I have a termination problem

> 4) I have a cabling problem

> ---
> [this gets even more annoying as I'm currently trying to debug why
> a new machine I just loaded with slackware's distribution keeps on
> giving me signal 11's when I try to compile the kernel (as well as
> when I run java :-(]

> please e-mail a reply

> --

1) Unlikely

2) Possible, but not terribly likely.

3) Possible.

4) The most likely.

I administer 4 Linux machines and occasionally work with 2 others (one's my home
machine, the others at work), all with 1542CF SCSI controllers.

I say from experience these controllers are HORRIBLY sensitive to cabling.

I recently added an external hard drive to my home machine (which allready had a SCSI
scanner connected externally and was working fine).  Immediately, I started getting SCSI
errors out the wazoo on all my devices, internal and external.  I had numerous files on
the external and internal hard disks get hosed.  Kernel panics due to screwed swap
partitions... you name it.  I double, tripple, and quadruple checked the terminations,
they were correct.

The problem:  I had bought a $16.00 SCSI cable to hook up the new external drive.
I replaced it with one that cost $50.00.  Same length, same *manufacturer*, and my
problems all went away.

So, my bet is on your cables.  If you are fortunate enough to be able to borrow
some SCSI cables, try swaping out your external cables, one by one, and see if you
can make the problem go away.  Then, replace the culprit cable.

Hope this helps,
--
------------------------------------------------------------------------

|  The Laitram Corporation     | http://www.gnofn.org/~marksu/         |
|  New Orleans, Louisiana      |                                       |
------------------------------------------------------------------------