[Bioclusters] Re: new on using clusters: problem running mpiblast (2)

[Zhiliang Hu]

Thanks Joe!

Here are my trials -- I am not sure exactly which does what so I tried 
with different combinations -- hopefully more informative.  Since the 
errors look "messy", for clarity I put them on a web page:

1.
> /opt/openmpi.gcc/bin/mpirun -np 3 -d 4 -mca btl tcp -machinefile 
machines /home/local/bin/mpiblast -p blastp -i ./bait.fasta -d ecoli.aa
Errors: http://www.animalgenome.org/~hu/share/tmp/test0.html

2.
>/opt/openmpi.gcc/bin/mpirun -np 3 -mca btl tcp -machinefile machines 
/home/local/bin/mpiblast -p blastp -i ./bait.fasta -d ecoli.aa
Errors: http://www.animalgenome.org/~hu/share/tmp/test1.html

3.
> /opt/openmpi.gcc/bin/mpirun -np 3 -d 4 -machinefile machines 
/home/local/bin/mpiblast -p blastp -i ./bait.fasta -d ecoli.aa
Errors: http://www.animalgenome.org/~hu/share/tmp/test2.html

Regards,
Zhiliang


On Fri, 21 Sep 2007, Joe Landman wrote:

> Date: Fri, 21 Sep 2007 11:20:10 -0400
> From: Joe Landman <landman at scalableinformatics.com>
> To: Zhiliang Hu <hu at animalgenome.org>
> Cc: HPC in Bioinformatics <bioclusters at bioinformatics.org>
> Subject: Re: [Bioclusters] Re: new on using clusters: problem running mpiblast
>      (2)
> 
> Zhiliang Hu wrote:
>> Thanks Joe!
>>
>> I tested with 'which' and 'whereis', they do find 'orted' on my system.
>> So I tried again with full path to "mpirun" (i am sorry I should have
>> done this earlier):
>>
>>> /opt/openmpi.gcc/bin/mpirun -np 3 -machinefile machines
>> /home/local/bin/mpiblast -p blastp -i ./bait.fasta -d ecoli.aa
>
> Lets have you add '-d 4' to the mpirun line in both cases.
>
> Also, I might suggest forcing it to use the ethernet if possible.  You
> would need to add in
>
> 	-mca btl tcp
>
> to make that happen.
>
>