[GE users] qdel not remove all instances of jobs

Reuti reuti at staff.uni-marburg.de
Fri Mar 9 07:10:58 GMT 2007


Simon,

Am 09.03.2007 um 01:01 schrieb Simon Gao:

> Reuti wrote:
>> Hi,
>>
>> Am 07.03.2007 um 23:03 schrieb Simon Gao:
>>
>>> I am having problems with qdel to completely remove jobs from a  
>>> compute node. After running "qdel <jobid>", the job spawned  
>>> processes continue to run even though the job has been removed  
>>> from queue.  What could cause such problem?
>>
>> are the processes jumping out of the processgroup? Are they  
>> started by a "&" in the jobscript? So, we need some more details  
>> about it. Maybe a "ps -e f -o pid,ppid,pgroup,command" from these  
>> surviving processes.
>>
>> -- Reuti
> Here is one example.
>
> Initially:
>
> [user1 at compute-0-39 ~]$ pstree pn | grep -A 1 user1
>        |-sshd(2875)---sshd(8263)---sshd(8265,user1)---tcsh(8266)-+- 
> pstree(8441)
>        |                                                             
> `-grep(8442)
> --
>        |-sge_execd(3170,sge)---sge_shepherd(8105)---perl 
> (8106,user1)-+-perl(8208)
>         
> |                                                                `- 
> sh(8227)---app1(8228)

may I beg you to post the relevant lines of:

ps -e f -o pid,ppid,pgrp,command

as it will provide more details. Thx - Reuti

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