[GE users] MPICH2 tight integration

skylar2 skylar2 at u.washington.edu
Fri Aug 14 01:15:31 BST 2009


reuti wrote:
> Am 13.08.2009 um 19:04 schrieb skylar2:
>
>   
>> No joy on the daemonless smpd front. SGE successfully starts up all  
>> the
>> processes, with one line like this per slot:
>>
>> /net/gs/vol3/software/sge/bin/lx24-amd64/qrsh -inherit sage001 env
>> PMI_RANK=0 PMI_SIZE=48 PMI_KVS=3B9DD9611EAE3B5240D7D4AE2DE
>> 3BDDE PMI_ROOT_HOST=sage001.grid.gs.washington.edu PMI_ROOT_PORT=60437
>> PMI_ROOT_LOCAL=0 PMI_APPNUM=0 /net/gs/vol3/software/mod
>> ules-sw-test/hpl-mpich2/2.0/Linux/RHEL5/x86_64/bin/ 
>> Linux_PII_CBLAS_gm/xhpl
>>     
>
> Is this a Pentium-II library?
>   


Yeah.
> The best thing would be to start with the mpihello program to get it  
> working with a simple program. Maybe there is some network address  
> problem or whatever. As I said: is it working with 2/4/8... slots on  
> 1/2/4... nodes with mpihello?
>
>   

That got me on the right track with smpd. I got mpihello working, and
then noticed that while I was using the MPICH2 compiled for smpd, HPL
was still looking for mpd. I pointed myself at HPL for smpd and life is
good now. I'm still not sure what's wrong with mpd, but I can live with
smpd for now.


Thanks!

-- 
-- Skylar Thompson (skylar2 at u.washington.edu)
-- Genome Sciences Department, System Administrator
-- Foege Building S048, (206)-685-7354
-- University of Washington School of Medicine

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=212201

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

    [ Part 2, "OpenPGP digital signature"  Application/PGP-SIGNATURE ]
    [ (Name: "signature.asc") 259 bytes. ]
    [ Unable to print this part. ]



More information about the gridengine-users mailing list