Sigh. It happened again. Some background info:
* Upgraded to Redhat 6.0, fresh install on a new machine.
* SCSI tape drive installed.
The tape drive will work ONCE. The next time the tape drive is accessed,
the following error message is given:
x kernel: st0: Error 2603007f (sugg. bt 0x20, driver bt 0x26, host bt
0x3).
x: last message repeated 4 times
[root@x /] mt -f /dev/st0 status
st0: Error 2603007f (sugg. bt 0x20, driver bt 0x26, host bt 0x3).
SCSI 2 tape drive:
File number=-1, block number=-1, partition=0.
Tape block size 0 bytes. Density code 0x0 (default).
Soft error count since last status=0
General status bits on (10000):
IM_REP_EN
We played it around with it the first time it happened and frustrated,
rebooted the machine. During the reboot of the machine, the modprobe
command which runs from rc.sysinit gave an error message saying that
"module st not found". This error message can be duplicated on the command
line.
After the reboot, the tape drive worked ONCE again, and then broke.
We've looked at everything we can think of. Some information gleaned from
the Internet and other posts:
* Redhat 6.0 doesn't come with the st.o kernel module
(True. This is not located on our drives at all).
* modprobe -c gives us "alias char-major-9 st" which matches up with some
information from the /proc system. "st" is located at 9.
* kerneld is not running.
* Someone suggested to rebuild the kernel with st.o support. Other
information I've seen says that the 6.0 kernel is already built with this
support, which is why the module is missing. One guy had luck rebuilding
the kernel... we don't have the inclination, the knowhow, or the
possibility of downtime to make this an option (Question: could we take the
st.o module from someone else's 2.25smp distribution and just plop it in
the right place?).
* Playing around with lsmod, insmod, and depmod -a didn't seem to prove
helpful, although more information using these commands can be forthcoming.
* I took a look back at temporarily unloading the scsi module (for the tape
drive) and then reloading it, but paranoia sets in when I assume that the
same scsi module is running the hard drives attached.
* One guy suggested to comment out the actual lines in rc.sysinit which
call the modprobe command if it fines anything related to st in the /proc
filesystem. We have not tried this, although I fail to see why this would
help - the commenting in the file makes it seem unimportant if modprobe
doesn't find the module.
* Another post from a newsgroup suggested that trying to open the actual
/dev/st0 or /dev/nst0 files would accomplish something - they only
accomplished getting the same error message from st0 as above.
* All commands are being run as root.
Any help would be appreciated. Or even suggestions. Hell, I'm desperate. I
do NOT want to reboot the machine everytime the tape drive breaks. That's
plain old crazy.
Kevin Hemenway
Total Net NH, LLC
--
To unsubscribe: mail [EMAIL PROTECTED] with "unsubscribe"
as the Subject.
