Hi all,

I did a search on the mailing list to try to find the best way to adjust 
slurmd’s OOM score so as not to place it in the line of fire in case of memory 
exhaustion. From what I can tell here, it and a few other system services have 
the default priority of 0:

2015-12-29T14:34:10-05:00 node75 kernel: [ pid ]   uid  tgid total_vm      rss 
cpu oom_adj oom_score_adj name
...
2015-12-29T14:34:10-05:00 node75 kernel: [21323]   493 21323    57393      445  
 7       0             0 munged
2015-12-29T14:34:10-05:00 node75 kernel: [21430]     0 21430    47074     4258  
 6       0             0 slurmd
2015-12-29T14:34:10-05:00 node75 kernel: [22031]     0 22031     1697      105  
16       0             0 mcelog

I searched the mailing list and only came up with someone suggesting SLURM be 
patched to set its priority higher. Searching the web, I found some indications 
that that patch might have been applied and that that functionality now exists 
by default. However, here I am. The system that prompts me to ask is running 
SLURM 14.06.3 on CentOS 6, but I am also running a system on CentOS 7 with 
SLURM 15.08.5 and I see that the slurmd has a score of 0 there too.

Is there a suggestion about how to best handle this these days?

--
____ *Note: UMDNJ is now Rutgers-Biomedical and Health Sciences*
|| \\UTGERS      |---------------------*O*---------------------
||_// Biomedical | Ryan Novosielski - Senior Technologist
|| \\ and Health | [email protected] - 973/972.0922 (2x0922)
||  \\  Sciences | OIRT/High Perf & Res Comp - MSB C630, Newark
     `'

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

Reply via email to