[torqueusers] Ulimit Help

[Chris Vaughan]

Thanks Rishi, 


We added " ulimit -u unlimited" to the pbs_mom start stop scripts and that has appeared to resolve the issues. 


Best Regards, 




Hi Chris, 
We faced the same problem with 2.3.6. It seemed like on pbs_mom startup at system boot did'nt inherited limits from /etc/security/limits.conf. We verified this by submitting a interactive job and listing the limits using ulimit -a. Subsequent restart of pbs_mom(w/o reboot) comes up with defined limits. We observed that ssh login for a user shows the correct limits for the node in question. Next what we tried was to make pbs_mom start after sshd, but the results were same. 


On Fri, Jun 18, 2010 at 6:35 PM, Chris Vaughan < chris at adaptivecomputing.com > wrote: 






Hi, 




I have a customer with the following error with ulimits for the job not being unlimited, does anyone know how to resolve this? Do I need to have this set when pbs_server and pbs_mom starts? 




Thanks, 




We've been having problems running Fluent software at our site. It works well when running interactively, but it give errors(see attached error file), when running it in a queue. You'll see in the file attached, it complains about 'locked down' memory, which it tries to use for MPI which is using locked memory for Infiniband communications. The /etc/security/limits.conf file on all nodes reads 

@ * hard memlock unlimited 
@ * soft memlock unlimited 



-- 





Chris Vaughan | Technical Consultant - EMEA 


Adaptive Computing | Mobile: +44 7800 973062 | Office: +44 1483 243578 


3000 Cathedral Hill | Guildford GU2 7YB | United Kingdom 






_______________________________________________ 
torqueusers mailing list 
torqueusers at supercluster.org 
http://www.supercluster.org/mailman/listinfo/torqueusers 




-- 
Regards-- 
Rishi Pathak 
National PARAM Supercomputing Facility 
Center for Development of Advanced Computing(C-DAC) 
Pune University Campus,Ganesh Khind Road 
Pune-Maharastra 



-- 





Chris Vaughan | Technical Consultant - EMEA 


Adaptive Computing | Mobile: +44 7800 973062 | Office: +44 1483 243578 


3000 Cathedral Hill | Guildford GU2 7YB | United Kingdom 





-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/torqueusers/attachments/20100621/6a958f75/attachment-0001.html