[GE users] Specifying max retry on a job?

Peter Bowmar pbowmar at yahoo.com
Sat Aug 26 13:22:14 BST 2006


    [ The following text is in the "iso-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

Hi Reuti,
    Thanks for the replies, I'll try out the env variable idea. I've
got the PDF versions of the docs, but like I said, they're pretty
sparse. I'd never have figured out your suggestion based on the docs!!
I guess I'm used to a very specialized render queue, whereas SGE is a
lot more general-purpose (and free) but I'm not against writing my own
wrappers and whatnot, I just need to figure out the hooks to do it.
    Thanks again! Very useful list, and some of my questions have been
answered by other's posts which is great.

Cheers,

Peter B

--- Reuti <reuti at staff.uni-marburg.de> wrote:

> Hi,
> 
> Am 26.08.2006 um 10:30 schrieb Peter Bowmar:
> 
> > Hi,
> >       Short time listener, first time poster :) I'm using SGE 6 to
> > render frames of animations, it works very well for that. I'm
> curious,
> > since I can't find it mentioned in the docs or the Qmon interface,
> how
> > do I specify a max limit on the number of times a job is re-run?
> >       If I get a bad frame for some reason that exits with error
> code
> > greater than 0, I exit with 99, which nicely triggers a re-run.
> > However, in some cases, it will just keep crashing (bug in  
> > renderer, or
> > my shaders or whatever) and I don't want it to keep re-running,
> > essentially plugging up the queue with failing tasks.
> 
> what about:
> 
> qalter -v COUNT=$value $JOB_ID
> 
> before the exit 99. The variable COUNT (or any you like) can be  
> tested in
> the script (nodes must be a submit host in this case).
> 
> >       Also, if there is a better source of docs than the ones
> shipped
> > with SGE (they're bare-bones to say the least :) please point me at
> > them. This list seems to be the definitive source of knowledge but
> I
> > don't want to keep posting questions that have answers elsewhere.
> 
> You can use the SUN docs:
> 
> http://gridengine.sunsource.net/documentation.html
> 
>   - Reuti
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
> For additional commands, e-mail: users-help at gridengine.sunsource.net
> 
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list