[GE users] Fairshare priority question

Chris Dagdigian dag at sonsorol.org
Tue Mar 8 14:50:55 GMT 2005


Hi Steve,

One thing that may help in debugging is to run the command:

# qconf -tsm

That will cause a one-time dump of interesting scheduling information to 
the location $SGE_ROOT/$SGE_CELL/common/schedd_runlog

For that particular scheduler interval the schedd_runlong messages 
should indicate the reasons for jobs and queues not being selected in 
that run. Running the -tsm a few times may be enlightening.

-Chris



slitster at rcn.com wrote:

> I may be missing something totally obvious here.
> 
> I've set up a functional share policy with 10000 tickets.
> 20 users have 100 tickets each.
> 
> The share policy doesn't seem to be working as it still 
> appears to be scheduling FIFO, ie. one user has 3000 jobs 
> queued, another user wants to launch 10 jobs, but they are 
> queued after the 3000 jobs with the same priority (0.55).
> 
> What have I missed?
> 
> Cheers,
> 
> Steve
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
> For additional commands, e-mail: users-help at gridengine.sunsource.net

-- 
Chris Dagdigian, <dag at sonsorol.org>
BioTeam  - Independent life science IT & informatics consulting
Office: 617-665-6088, Mobile: 617-877-5498, Fax: 425-699-0193
PGP KeyID: 83D4310E iChat/AIM: bioteamdag  Web: http://bioteam.net

---------------------------------------------------------------------
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