/dev/rmt0.1 weirdness (RS/6k, AIX 3.1.3002)

[Robert E. Wargaski Jr.]

I have been receiving some odd errors from our RS/6k when it runs our
incremental backup script.  The script is as follows (host1 is the
RS/6k, and host2 is the Sun whose /usr1 disk is being backed up to the
RS/6k tape unit):

   #!/bin/sh
   tctl rewind #make sure tape rewound
   /etc/umount /u #unmount /u for backing up
   /etc/backup -f /dev/rmt0.1 -9 -u /u #actually back up /u---no rewind
   /etc/mount /u #remount /u
   rsh host2 /etc/dump 9ubf 500 host1:/dev/rmt0.1 /usr1
	#backup host2 to host1's tape drive		
   tctl rewind #rewind tape

This is the tail end of the stdout from the backup script (set up in
root's crontab to run at 4:30 every morning):

   DUMP: DUMP: 30281 tape blocks on 1 tape(s)
   DUMP: DUMP IS DONE
   DUMP: level 9 dump on Thu Mar  7 04:32:58 1991
   DUMP: write: There is a request to a device or address that does not exist.

   DUMP: Write error on tape 1
   DUMP: fopen on /dev/tty fails
   DUMP: Bad return code from dump: 65280
   /dev/rmt0.1: Cannot create another process at this time.

The lines prefaced with DUMP: are stdout/stderr from the dump program
on the Sun.  It seems that the dump completes, but then chokes when
it's wrapping things up.

This script has been running since September with no problems.  The
odd thing about the AIX error at the end (about /dev/rmt0.1 not being
able to create another process) is that the script is run at 4:30 in
the morning when nobody is logged on . . .  Does anyone know why
/dev/rmt0.1 can't create another process?

Thanks.

Rob
-- 

Robert E. Wargaski Jr.			 This is stupid. -- Vila
wargaski at nwu.edu			 When did that ever stop us. -- Avon
ACNS DSS, Northwestern University			Moooo!