From: Craig Allison <craig@(email surpressed)>
Subject: rush: 'rush -submit': No available jobids!
   Date: Mon, 07 Jul 2008 07:04:43 -0400

Hi guys

I wonder if someone can help, I'm getting the above error on a  
particular submit host, same job works on other hosts...

How do I "reset" the jobids for this box?  I've cleared down all the  
old jobs but still get the issue.

Cheers

Craig

Craig Allison
Digital Systems & Data Manager
The Senate Visual Effects
Twickenham Film Studios
St.Margarets
Twickenham
TW1 2AW

www.senatevfx.com

Craig Allison wrote:
> I wonder if someone can help, I'm getting the above error on a
> particular submit host, same job works on other hosts...

	If you can't solve it, as a short term solution try setting
	the 'Submit Host:' prompt to point at some other machine
	so the user at that machine can submit normally.

	The jobid for their jobs will simply be on some other host,
	but they'll still be able to manange their job.

> How do I "reset" the jobids for this box?  I've cleared down all the old
> jobs but still get the issue.

	Hmm, dumping old jobs should free up the jobids..
	are you sure the jobs are actually dumping?

	Try including in an email directly to me the output of
	these commands run on this machine:

If Windows:
rush -lj > togreg.txt
rush -dcatlog >> togreg.txt
dir \rush\var >> togreg.txt
cacls \rush\var\* >> togreg.txt
cacls \rush\var   >> togreg.txt

If Unix:
rush -lj > togreg.txt
rush -dcatlog >> togreg.txt
ls -la /usr/local/rush/var >> togreg.txt
ls -lad /usr/local/rush/var >> togreg.txt

	If windows, it's possible it might be a permission problem
	on one of the files in the rush/var directory. Try running
	the following as administrator:

cacls c:\rush\var   /E /C /G everyone:F
cacls c:\rush\var\* /E /C /G everyone:F

	..or if unix, it might be the perms on the rush or rushd
	executable that got messed up by someone with a runaway 'chown'
	or 'chmod' command:

chown 0:0  /usr/local/rush/bin/{rush,rushd}
chmod 4755 /usr/local/rush/bin/{rush,rushd}

	..then restart the daemon on that machine.

	If you want to just clear out all the jobs, stop the daemon
	and rename out the rush/var/jobs-checkpoint file to something
	else, then start the daemon. Then any jobs should be gone,
	but only do that as the last step.


-- 
Greg Ercolano, erco@(email surpressed)
Seriss Corporation
Rush Render Queue, http://seriss.com/rush/
Tel: (Tel# suppressed)
Fax: (Tel# suppressed)
Cel: (Tel# suppressed)

Greg Ercolano wrote:
> Craig Allison wrote:
>> I wonder if someone can help, I'm getting the above error on a
>> particular submit host, same job works on other hosts...

	This was solved quickly in private email.

	Apparently users weren't dumping jobs, and they had been accumulating
	over the last few months.

	Seems there were in fact 1000 jobs on the server in question, but
	because there were so many, they weren't showing up in 'All Jobs'
	(probably due to a "NO RESPONSE FROM.." error that went unseen),
	making it seem like there were no jobs.

	The jobs did show up by putting the job server's hostname followed
	by ".0" in the "Jobid:" prompt of irush, and hitting 'Jobs' to see
	all 1000 jobs. Or by just running 'rush -lj' on the machine in question.