[GE users] Many "Job ... Continued" Messages

Reuti reuti at staff.uni-marburg.de
Tue Dec 13 15:46:16 GMT 2005


Hi Jonathan,

Am 13.12.2005 um 16:24 schrieb Jonathan T Zink:

>
> All,
>
> First, thank you for the group's help in solving another problem I  
> had with GridEngine.  This community is clearly committed to  
> helping others...
>
> I have a (perhaps) interesting problem:
>
> I have been submitting jobs with qsub option "-m aes". Each night  
> the same set of tests are submitted -- this has been working  
> successfully for several months.
>
> However, I am now receiving 10k+ emails nightly of the form "Job  
> <id#> (<job name>) Continued".   These occur about 2/minute  
> (irregular both by job and time) and are associated with 5 to 10  
> different jobs at a time (there are >100 jobs running).  It doesn't  
> appear to be the application because the night-to-night emails are  
> associated with different submitted jobs and most of the jobs run  
> without issue.  It also doesn't appear to be from any particular  
> queue or machine -- >20 machines have been affected from 5  
> different queues.
>
> Any thoughts on the problem?  Network? Time sync?  The IT guys have  
> assured me that they have not made any changes lately....

this might be the result of issue:

http://gridengine.sunsource.net/issues/show_bug.cgi?id=1665

, and as you specified "s(uspend)" in the -m option, you get all the  
emails. It's fixed in u7, not to send the signals endlessly, and so  
you should only get one email for each suspend/continue.

HTH - Reuti


> Thanks in advance for your help,
> Jonathan
>
> Jonathan Zink
> Raytheon Missile Systems
> 520.794.0242
> 520.489.0742 pager
> jtzink at raytheon.com


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