Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

Subject: [OMPI devel] MTT: added long COMM_SPAWN test
From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2009-05-20 13:27:36

I just added a loop_spawn test to the IBM test suite that,
unsurprisingly, does a big ol' loop over COMM_SPAWN (2000 spawns, to
be exact).

Those who are running the IBM test suite in MTT, take note: it takes a
long time to complete -- probably longer than your default timeout. I
added the following to my INI file in my IBM test run section:

# Big loop o' spawns
loop_spawn:tests = dynamic/loop_spawn
loop_spawn:np = 1
loop_spawn:pass = &and(&test_wifexited(), &eq(&test_wexitstatus(),0))
loop_spawn:exclusive = 1
loop_spawn:timeout = 600

On the trunk with btl=tcp,self and np=1 on 3-year-old xeon 4-core
server, it ran in about 3.5 minutes. It failed with btl-openib,self
after about 1300 spawns (claiming not enough qp's). I'll file a
ticket about that.

It fails on v1.3 after ~120 spawns (regardless of BTL) claiming it had
run out of fd's. I'll file a ticket about that.

Jeff Squyres
Cisco Systems