Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!cbatt!ucbvax!CHEETA.ISI.EDU!rod
From: rod@CHEETA.ISI.EDU.UUCP
Newsgroups: mod.computers.vax
Subject: More on Queue Deaths
Message-ID: <870303151007.00h@Cheeta.ISI.Edu>
Date: Tue, 3-Mar-87 18:14:59 EST
Article-I.D.: Cheeta.870303151007.00h
Posted: Tue Mar  3 18:14:59 1987
Date-Received: Fri, 6-Mar-87 02:13:44 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 25
Approved: info-vax@sri-kl.arpa


	In my earlier note I said that our printer symbiont was dying
with something about privileged images not installed. I'm embarassed;
it wasn't the symbiont that died that way, it was ANALIMDMP's attempt
to put me in the debugger! I was flying without documentation and didn't
realize it was trying to put me in the debugger.

	It seems ANALIMDMP wants Sys$Library:UISSHR to be installed.
Why did it come installed on my microvax, but not on our cluster?

	Well, now it's time to try and figure out why things die. If
the image associated with a queue bombs, should that corrupt the
queue structure? Our biggest problems have been with our printer
symbiont, but the queue manager dies at the same time. Sometimes
simply restarting the queue manager and all the queues clears things
up, sometimes job control gets out of whack, too, with errors
reading job control mailbox.

    I didn't write our symbiont; someone who is no longer here did.
I've got a feeling it's sending something bogus to job control and
really screwing things up; how vulnerable is job control to others'
stupidity?

	--Rod
	  rod@ISI.Edu