--On Wednesday, December 03, 2003 14:40:03 -0500 Josiah Ritchie <jritchie AT 
bible DOT edu> wrote:

> Frank Smith scripted ::
> 
>> --On Wednesday, December 03, 2003 10:54:20 -0500 Josiah Ritchie
>> <jritchie AT bible DOT edu> wrote:
>> 
>>> So I'm running these commands that Jay gave me.
>>> 
>>># mt -f /dev/nst0 rewind
>>># mt -f /dev/nst0 status
>>># tar cbf 64 /dev/nst0 /etc
>>># tar cbf 64 /dev/nst0 /var
>>> 
>>> Then I ran:
>>># tar cbf 64 /dev/nst0 /usr
>>> and received this message.
>>> 
>>> tar: Removing leading `/' from member names
>>> tar: /dev/nst0: Wrote only 0 of 32768 bytes
>>> tar: Error is not recoverable: exiting now
>>> 
>>> Does that indicate that the tape is full? I'm figuring not, because it said
>> it> didn't write anything from that last command.
>>> 
>>> When I entered the next command:
>>># mt -f /dev/nst0 rewind
>>> /dev/nst0: No such device or address
>> 
>> If the mt rewind and status commands you listed at the top worked, but
>> now you get an error, I would guess the drive or interface has locked up.
>> Check your system logs for errors.
> 
> System logs report nothing about the tape drive. What is the service I should 
> be
> looking for? I'm relatively sure it isn't there anyway because my logs both
> /var/log/messages and syslog have only [sn]mbd and rpc.* services reporting.
> Nothing from the kernel or elsewhere. I finally broke down and rmmod/modprobe
> both st and aic7xxx. This resulted in no change of anything. I might just need
> to do a full powerdown of the device and cut the power to it to reset it 
> unless
> anyone has a better idea.

The log entries would reference the st driver and/or the SCSI interface
driver (aic7xxx).  You might wnat to check the other logs in /var/log
or run dmesg to see if it is still in the kernel's log buffer.
   If the tape drive is the only device on that SCSI chain you can safely
cycle power on the tape drive without rebooting the machine. Don't do it
if there is a disk on  the same interface or you might corrupt your disk
(and read the manual, since some controllers have internal and external
connectors for the same interface). If all else fails, power it all off
and start over (generally the power-off sequence is from the controller
out, and power-on is from the last device in).
   Since the tape device worked for a short while, I would suspect problems
with SCSI termination (either doubly terminated, or not at all).

Frank

> 
> Thanks,
> Josiah



-- 
Frank Smith                                      fsmith AT hoovers DOT com
Systems Administrator                           Voice: 512-374-4673
Hoover's Online                                   Fax: 512-374-4501