Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

Subject: Re: [OMPI devel] System V Shared Memory for Open MPI:Request forCommunity Input and Testing
From: N.M. Maclaren (nmm1_at_[hidden])
Date: 2010-05-04 11:07:47


On May 4 2010, Jeff Squyres wrote:

> If there's a sleep(1) in the run-time test, that would be an annoying
> source of delay in the startup of a job. This is not a deal-breaker, but
> it would be nice(r) if there was a "fast" run-time check that could be
> checked during the sysv selection logic (i.e., sysv could disqualify
> itself if the feature is not available at runtime). Keep in mind that the
> run-time check will be run in parallel across the whole job, so it's
> (more or less) a constant amount of time that is added to job startup.

Yes. I was thinking more of a configuration test and would recommend a
higher value - sleep(1) is a bit unpredictable as to whether it returns
almost immediately.

In the absence of a proper specification of such facilities, it really
comes down to which is the least unsatisfactory piece of guesswork :-(

Regards,
Nick Maclaren.