Timeout on 2842VLB during backup to ftape

[Bob Wheater]

Hi:
	I was unable to get an answer from the redhat.list so
	I'm asking here.

	I'm getting a timeout on the adaptec 2842vlb went doing a
	backup to ftape (colorado jumbo 250). This happens went
	using tar to backup a IBM SCSI 2.1GB (MODEL DORS-32160).
	The partition is approximately 440 megs in size.
        
	The kernel that I am using is 2.0.32. However, I noticed
	someone else in newsgroup comp.periphs.scsi has had a timeout
	related to ftape on a different version of the kernel.

	IS THERE A FIX FOR THIS PROBLEM OR WILL THERE BE ONE?

	thanks!
	regard,
	Bob Wheater

-------- excerpt from /var/log/messages ----------------
Jul 18 17:58:58 localhost kernel: ftape-2.08 960314
Jul 18 17:58:58 localhost kernel:  (c) 1993-1995 Bas Laarhoven
(bas at vimec.nl)
Jul 18 17:58:58 localhost kernel:  (c) 1995-1996 Kai Harrekilde-Petersen
(khp at dolphinics.no)
Jul 18 17:58:58 localhost kernel:  QIC-117 driver for
QIC-40/80/3010/3020 tape drives
Jul 18 17:58:58 localhost kernel:  Compiled for kernel version 2.0.32
with versioned symbols
Jul 18 17:58:58 localhost kernel: [000]    fdc-io.c (fdc_config) - Using
the standard fdc controller.
Jul 18 17:58:58 localhost kernel: [001]    fdc-io.c (fdc_probe) - Type
82077AA FDC found.
Jul 18 17:58:59 localhost kernel: [002]   ftape-ctl.c
(ftape_activate_drive) - drive wakeup method: Colorado.
Jul 18 17:59:02 localhost kernel: [003]    ftape-io.c
(ftape_set_data_rate) - post QIC-117B drive @ 1 Mb/s.
Jul 18 17:59:02 localhost kernel: [004]    ftape-io.c
(ftape_set_data_rate) - 307 ft. QIC-80 tape.
Jul 18 17:59:03 localhost kernel: [005]   ftape-read.c
(read_header_segment) - header segments are 0 and 1.
Jul 18 17:59:03 localhost kernel: [006]    ftape-bsm.c
(extract_bad_sector_map) - 30 bad sectors.
Jul 18 18:13:37 localhost kernel: [007]      ftape-rw.c
(ftape_dumb_stop) - aborting fdc operation.
Jul 18 18:18:47 localhost kernel: scsi : aborting command due to timeout
: pid 96816, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 24 ed
42 00 00 04 00
Jul 18 18:18:47 localhost kernel: (scsi0:1:0) Abort_reset, scb flags
0x1, Data-In phase, SCSISIGI 0x44, SEQADDR 0x125, SSTAT0 0x0,
SSTAT1 0x3
Jul 18 18:18:47 localhost kernel: scsi0: abort message in message buffer
Jul 18 18:18:47 localhost kernel: scsi : aborting command due to timeout
: pid 96817, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 26 43
d6 00 00 02 00
Jul 18 18:18:50 localhost kernel: SCSI host 0 abort (pid 96816) timed
out - resetting
Jul 18 18:18:51 localhost kernel: SCSI bus is being reset for host 0
channel 0.
Jul 18 18:18:52 localhost kernel: (scsi0:-1:0) Reset device, active_scb
1
Jul 18 18:18:52 localhost kernel: scsi0: (targ -1/chan A) matching scb
to (targ 1/chan A)
Jul 18 18:18:52 localhost last message repeated 3 times
Jul 18 18:18:52 localhost kernel: scsi0: Resetting current channel A
Jul 18 18:18:52 localhost kernel: scsi0: Channel reset, sequencer
restarted
Jul 18 18:18:52 localhost kernel: (scsi0:1:0) Aborting scb 0
Jul 18 18:18:52 localhost kernel: (scsi0:1:0) Aborting scb 1
Jul 18 18:54:04 localhost init: Switching to runlevel: 6

--------similar problem reported in newsgroup -------------------
Subject:      [Q] How to track down random error?
From:         phi at bh.marintek.sintef.no (Per Henning Isaksen)
Date:         1998/07/27
Message-ID:   <901547611.900883 at halvan.trd.sintef.no>
Newsgroups:   comp.periphs.scsi 
[More Headers]
[Subscribe to comp.periphs.scsi]


I have problems with timeout (sometimes!), problem may
block access to partition completely.

System is Adaptec2940UW, Quantum Fireball 3.2GB Ultra SCSI-2,
Tyan Tomcat MB and Linux 2.0.34.

First impression is that it is a thermal problem, i.e. it
disappears after the PC has been off for some time, cold
boot is not sufficient. However, heating disk/controller
reasonably (?) does not provoke any error.

Second impression is that it only appears when running
backup (ftape floppy driver). Coincidence ??

Is there any  software available that could be used
to test MB/controller/disk when the problem decides to
show up?

Any comments appreciated!

pH

PerHenning.Isaksen at marintek.sintef.no


To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message