[GE users] MVAPICH 0.9.9 and SGE 6.2u1

reuti reuti at staff.uni-marburg.de
Tue Aug 4 16:54:13 BST 2009


    [ The following text is in the "utf-8" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some characters may be displayed incorrectly. ]

Hi,

Am 03.08.2009 um 23:36 schrieb rgigon:

> Thanks so much for your response.
>
> Can you point me to the correct HOWTO on how to apply this patch?   
> The HOWTO I used was:  http://gridengine.sunsource.net/howto/ 
> mvapich/MVAPICH_Integration.html  which did not mention a patch or  
> recompiling MVAPICH.
>
> I found the patch you mentioned below attached to a message thread,  
> but upon examination it looks like it is specific to  
> mvapich-0.9.5.117.  I'm using mvapich-0.9.9.  Do I just need to  
> edit the file to reflect the directory names for 0.9.9 or is there  
> a different patch required.

the Howto seems to be some years old. The patch is also in the  
archive of this HowTo.

I don't know about any newer one. There was already a similar  
discussion on the mailing list when you search for "mvapich". I don't  
know whether it's of any help. The usually approach for a Tight  
Integration is to replace any hard-coded "/usr/bin/rsh" to a plain  
"rsh" so that SGE's rsh-wrapper can catch it and route it to SGE's qrsh.

-- Reuti


> Many thanks for your help!
>
> Best regards,
> Roberta
>
> ---------------------------------------------------------------------- 
> -----------------------
> Roberta M. Gigon
> Schlumberger-Doll Research
> One Hampshire Street, MD-B253
> Cambridge, MA 02139
> 617.768.2099 - phone
> 617.768.2381 - fax
>
> This message is considered Schlumberger CONFIDENTIAL.  Please treat  
> the information contained herein accordingly.
>
> -----Original Message-----
> From: reuti [mailto:reuti at staff.uni-marburg.de]
> Sent: Saturday, August 01, 2009 5:09 PM
> To: users at gridengine.sunsource.net
> Subject: Re: [GE users] MVAPICH 0.9.9 and SGE 6.2u1
>
> Hi,
>
> Am 29.07.2009 um 18:32 schrieb rgigon:
>
>> Hi there,
>> I?m trying to achieve tight integration between MVAPICH 0.9.9 and
>> SGE 6.2u1
>> I followed the directions in the HOWTO and am able to submit jobs,
>> but when I use qdel, it still does not kill all the processes.
>>
>> Results from ps on master:
>>
>> 6553  4370  6553  \_ sge_shepherd-1302 -bg
>>  6578  6553  6578      \_ -bash /opt/sge/default/spool/fire7/
>> job_scripts/1302
>>  6669  6578  6578          \_ /usr/local/mpi/intel/mvapich-0.9.9/
>> bin/mpirun_rsh -rsh -np 4 -hostfile /tmp/1302.1.normal.q/machines ./
>> testtight.x
>>  6670  6669  6578              \_ /usr/bin/rsh fire7 cd /people4/
>> jliu17/roberta/source; /usr/bin/env LD_LIBRARY_PATH=/usr/local/mpi/
>> intel/mvapich-0.9.9/lib/shared:/usr/local/mpi/intel/mvapich-0.9.9/
>> lib:/usr/local/mpi/intel/mvapich-0.9.9/lib/shared: MPIRUN_MPD=0
>> MPIRUN_HOST=fire7 MPIRUN_PORT=60546 MPIRUN_RANK=0 MPIRUN_NPROCS=4
>> MPIRUN_ID=6669    NOT_USE_TOTALVIEW=1  ./testtight.x
>>  6678  6670  6578              |   \_ [rsh] <defunct>
>>  6671  6669  6578              \_ /usr/bin/rsh fire7 cd /people4/
>> jliu17/roberta/source; /usr/bin/env LD_LIBRARY_PATH=/usr/local/mpi/
>> intel/mvapich-0.9.9/lib/shared:/usr/local/mpi/intel/mvapich-0.9.9/
>> lib:/usr/local/mpi/intel/mvapich-0.9.9/lib/shared: MPIRUN_MPD=0
>> MPIRUN_HOST=fire7 MPIRUN_PORT=60546 MPIRUN_RANK=1 MPIRUN_NPROCS=4
>> MPIRUN_ID=6669    NOT_USE_TOTALVIEW=1  ./testtight.x
>>  6680  6671  6578              |   \_ [rsh] <defunct>
>>  6672  6669  6578              \_ /usr/bin/rsh fire8 cd /people4/
>> jliu17/roberta/source; /usr/bin/env LD_LIBRARY_PATH=/usr/local/mpi/
>> intel/mvapich-0.9.9/lib/shared:/usr/local/mpi/intel/mvapich-0.9.9/
>> lib:/usr/local/mpi/intel/mvapich-0.9.9/lib/shared: MPIRUN_MPD=0
>> MPIRUN_HOST=fire7 MPIRUN_PORT=60546 MPIRUN_RANK=2 MPIRUN_NPROCS=4
>> MPIRUN_ID=6669    NOT_USE_TOTALVIEW=1  ./testtight.x
>>  6676  6672  6578              |   \_ [rsh] <defunct>
>>  6673  6669  6578              \_ /usr/bin/rsh fire8 cd /people4/
>> jliu17/roberta/source; /usr/bin/env LD_LIBRARY_PATH=/usr/local/mpi/
>> intel/mvapich-0.9.9/lib/shared:/usr/local/mpi/intel/mvapich-0.9.9/
>> lib:/usr/local/mpi/intel/mvapich-0.9.9/lib/shared: MPIRUN_MPD=0
>> MPIRUN_HOST=fire7 MPIRUN_PORT=60546 MPIRUN_RANK=3 MPIRUN_NPROCS=4
>> MPIRUN_ID=6669    NOT_USE_TOTALVIEW=1  ./testtight.x
>>  6683  6673  6578                  \_ [rsh] <defunct>
>
> a Tight Integration needs "rsh" compiled in, not an absolute path
> like "/usr/bin/rsh" which shows up above. Did you also recompile
> MVAPICH after applying the "mvapich-rsh.patch" from the HowTo archive?
>
> -- Reuti
>
>
>> Your thoughts are greatly appreciated.
>>
>> Best regards,
>> Roberta
>>
>> --------------------------------------------------------------------- 
>> -
>> -----------------------
>> Roberta M. Gigon
>> Schlumberger-Doll Research
>> One Hampshire Street, MD-B253
>> Cambridge, MA 02139
>> 617.768.2099 - phone
>> 617.768.2381 - fax
>>
>> This message is considered Schlumberger CONFIDENTIAL.  Please treat
>> the information contained herein accordingly.
>>
>
> ------------------------------------------------------
> http://gridengine.sunsource.net/ds/viewMessage.do? 
> dsForumId=38&dsMessageId=210598
>
> To unsubscribe from this discussion, e-mail: [users- 
> unsubscribe at gridengine.sunsource.net].
>
> ------------------------------------------------------
> http://gridengine.sunsource.net/ds/viewMessage.do? 
> dsForumId=38&dsMessageId=210803
>
> To unsubscribe from this discussion, e-mail: [users- 
> unsubscribe at gridengine.sunsource.net].
>

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=210912

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list