[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Condor-users] Only 40 condor_shadows running on windows



Hi All

I am submitting 2040 jobs to a 900+ node windows condor pool and for the life
of me I can't work out why I am only getting a maximum of 40 condor_shadows
running concurrently on my submit host.  BTW condor_q reports that currently 270
jobs are running, 1766 are idle, and 4 have completed :) - How can 270 run
concurrently when there are only 40 shadows running ?

It's not lack of RAM (2G) or CPU (Dual Xeon 3.2) or Network (100 MBit).  It's
not my condor config file as MAX_JOBS_RUNNING is set at 20000 and RESERVED_SWAP
is commented out so in theory condor should soak up everything available with
condor_shadows up to 20000 of them.  

Does anybody know if there a limitation in condor for windows or in windows
itself re running more than 40 condor_shadow processes concurrently.  Does each
condor_shadow manage more than one remote job at once? BTW Each job lasts about
an hour and a half so it's not that the jobs are completing fast.

I hope somebody can clear up this conundrum...

Thanks in advance

James

---

James Osborne
High End Computing Support Engineer
Information Services, Cardiff University, 50 Park Place, Cardiff CF10 3AT,
Wales UK
Tel: +44(0)29 2087 4657
Fax: +44(0)29 2087 4285
Email: osborneja1@xxxxxxxxxxxxx 
Web: http://www.cardiff.ac.uk/insrv/condor