[mvapich-discuss] Diagnosing Slow MVAPICH2 Startup
Thompson, Matt (GSFC-610.1)[SCIENCE SYSTEMS AND APPLICATIONS INC]
matthew.thompson at nasa.gov
Tue Jan 12 10:05:24 EST 2016
On 01/12/2016 10:03 AM, Panda, Dhabaleswar wrote:
> Hi Matt,
>
> Thanks for your note. We will take a look at the ZiaTest to see what could be going-on here. At a first glance,
> you should not run MVAPICH2 with MV2_USE_SHMEM_COLL=0? There are many optimized collectives which are disabled by making this parameter=0. Have you run the test without setting this parameter=0? Just use the
> default version with affinity=0.
>
DK,
Yes, the 2.2b_bb is just MV2_ENABLE_AFFINITY=0. I only tried the other
one because someone else had reported a startup slowdown with it. It was
dartboard-throwing.
Matt
PS: I really hope this goes through well. On my box, the last messages I
sent came back as some base64 nonsense with no subject. Maybe my mailing
list subscription is screwed up...?
--
Matt Thompson, SSAI, Sr Scientific Programmer/Analyst
NASA GSFC, Global Modeling and Assimilation Office
Code 610.1, 8800 Greenbelt Rd, Greenbelt, MD 20771
Phone: 301-614-6712 Fax: 301-614-6246
http://science.gsfc.nasa.gov/sed/bio/matthew.thompson
More information about the mvapich-discuss
mailing list