Hi,
Le 09/03/2017 à 23:21, Reuti a écrit :
Hi,
Am 09.03.2017 um 17:20 schrieb Jerome Poitout:
> Hello,
> OGS/GE 2011.11p1
> I have an issue while submitting numerous jobs in a short
time (over 300
> - not so much for me...) with -sync y option. It seems that
qmaster
> cannot handle all the requests and i get huge load on the
head server
> (>400) and memory gets almost full (32GB).
I have no experience with such a huge amount of hanging -sync
jobs, but it doesn't sound to me like a problem of SGE but of the
hardware.
Besides the memory, what type of CPU is running the qmaster?
model name : Intel(R) Xeon(R) CPU E5-2430 v2 @ 2.50GHz (6 Cores,
12 Threads)
> These jobs are run by a third party product that does not
support job
> arrays (as far as we currently know).
Sometimes it's possible to use a wrapper which catches the `qrsh
-sync …` calls and route it to another type of workflow. E.g.
catching these calls and write them into one file and return
immediately (without any computation done so far, it's outside of
SGE). When all computation requests are written to that file, this
file can then be submitted by a usual `qsub`, even as a array job
with some lines around so that array index i handles the i-th
line's request. Depends on the particular application.
I'll forward this info
-- Reuti
> Then I get some timeout while trying to qstat something...
> [root@ ~]# qstat -u user
> error: failed receiving gdi request response for mid=1 (got
syncron
> message receive timeout error).
> Any idea on how to raise the number a jobs that can be qsub
in a short
> time ? I am almost sure that a qmaster params can be used but
as I am in
> production environment, I prefer to be careful...
> Best regards,
> Jérôme
> _______________________________________________
> users mailing list
> users@gridengine.org
> https://gridengine.org/mailman/listinfo/users
>
|
signature.asc
Description: OpenPGP digital signature
_______________________________________________
users mailing list
users@gridengine.org
https://gridengine.org/mailman/listinfo/users
