[Bioclusters] GCG Winsconsin Package and GridEngine

[Joe Landman]

If you wanted to find out what the bottleneck with srsupdate, you could
instrument the machine on which it is running to examine its state
before during and after a run.

Srsupdate does an index regeneration as I remember, and this is (as
database operations go) not a fast thing.  It would be hard to
parallelize unless your underlyiing database allowed parallelization.

My guess is (without any real data analysis or machine observation to
back it up) that your database machine or database system is slow (or is
crunching on a huge database, or has a complex set of indices to
calculate).

Joe

On Thu, 2002-05-09 at 06:24, Andrew Hynes wrote:
> Hi All,
> 
> Does anyone have any experience of distributing GCG applications (e.g.
> pileup and srsupdate) using Sun's GridEngine?  After a trawl through the web
> it seems that some GCG applications FrameSearch, BLAST, FASTA, FASTAX,
> FASTX, TFASTA, TFASTX and SSearch) are threaded and can take advantage of a
> multi-processor architecture.  I am more interested in distributing
> cpu-intensive jobs like pileup and srsupdate that are not threaded.  For
> pileup I suppose the way to go is to split up the input file and kick off a
> number of jobs on different servers using GridEnginec - does anyone have any
> comments?  As for srsupdate, I'm not sure how to procede, but its pig slow
> and I am sure there must be a way to speed it up with distribution.
> 
> Thanks,
> 
> Andrew
> 
> Andrew M Hynes PhD
> Bioinformatician
> Software Engineering
> Oxford GlycoSciences
> The Forum
> 86 Milton Park
> Abingdon
> Oxon OX14 4RY
> 01235 208065
> 
> 
> **********************************************************************
> The information transmitted by this email is private and
> confidential and is intended for the use of the intended 
> recipients specified therein.
> If you are neither an intended recipient nor an employee 
> or agent responsible for delivery to an intended recipient, 
> you should be aware that any dissemination, distribution 
> or copying of this communication is strictly prohibited.
> If you received this communication in error, please 
> notify us immediately.
> **********************************************************************
> _______________________________________________
> Bioclusters maillist  -  Bioclusters@bioinformatics.org
> http://bioinformatics.org/mailman/listinfo/bioclusters