Comments on: Why MPI is Good for You (part 2) http://blogs.cisco.com/performance/why-mpi-is-good-for-you-part-2/ Tue, 18 Jun 2013 04:08:32 +0000 hourly 1 By: More Reasons Why MPI is Good for You | insideHPC.com http://blogs.cisco.com/performance/why-mpi-is-good-for-you-part-2/#comment-688261 More Reasons Why MPI is Good for You | insideHPC.com Tue, 30 Oct 2012 15:12:25 +0000 http://blogs.cisco.com/?p=88653#comment-688261 [...] at the MPI Blog, Cisco’s Jeff Squyres writes that recent bug-driven crashes remind that implementing complex [...]

   0 likes

]]> By: Jeff Squyres http://blogs.cisco.com/performance/why-mpi-is-good-for-you-part-2/#comment-687825 Jeff Squyres Mon, 29 Oct 2012 16:46:50 +0000 http://blogs.cisco.com/?p=88653#comment-687825 John Hearns had a good reply about OOM killer on the Open MPI user’s mailing list:

http://www.open-mpi.org/community/lists/users/2012/10/20577.php

   0 likes

]]> By: Jeff Squyres http://blogs.cisco.com/performance/why-mpi-is-good-for-you-part-2/#comment-687707 Jeff Squyres Mon, 29 Oct 2012 12:32:46 +0000 http://blogs.cisco.com/?p=88653#comment-687707 Good point.

But keep in mind that it only takes one errant / buggy user space application to trigger the OOM killer. So “should almost never trigger” assumes that you have well-behaved users who don’t accidentally malloc() into oblivion. :-)

   0 likes

]]> By: Phil Miller http://blogs.cisco.com/performance/why-mpi-is-good-for-you-part-2/#comment-687503 Phil Miller Mon, 29 Oct 2012 02:36:02 +0000 http://blogs.cisco.com/?p=88653#comment-687503 It seems the right solution to that sort of issue is not suggesting disabling the OOM killer, which should almost never trigger, except as a last resort, but rather configuring the whole system properly. Put your management processes in a separate cgroup from your application processes, and prioritize the latter for OOM to target.

   1 like

]]>