Re: [OMPI devel] 1.6.1rc1 posted

2012-08-18 Thread Yevgeny Kliteynik
Hi Jeff, On 8/18/2012 3:54 AM, Jeff Squyres wrote: > Mike / Yevgeny -- > > Can you comment on what is going on here? It would be really good to > understand exactly what these 2 MLX4 parameters are (e.g., why you suggested > increasing one and not the other), and why there would be differences

Re: [OMPI devel] 1.6.1rc1 posted

2012-08-17 Thread Jeff Squyres
Mike / Yevgeny -- Can you comment on what is going on here? It would be really good to understand exactly what these 2 MLX4 parameters are (e.g., why you suggested increasing one and not the other), and why there would be differences in registering small numbers of large chunks of contiguous m

Re: [OMPI devel] 1.6.1rc1 posted

2012-08-14 Thread Paul Kapinos
Hi Jeff, Hi All, On 08/07/12 18:51, Jeff Squyres wrote: So I'm not 100% clear on what you mean here: when you set the OFED params to allow registration of more memory than you have physically, does the problem go away? We are talking about machines with 24GB RAM (S) and 96GB RAM (L). The defa

Re: [OMPI devel] 1.6.1rc1 posted

2012-08-07 Thread Jeff Squyres
On Aug 3, 2012, at 6:24 PM, Paul Kapinos wrote: > testing our well-known example of the registered memory problem (see > http://www.open-mpi.org/community/lists/users/2012/02/18565.php) on > freshly-installed 1.6.1rc2, found out that "Fall back to send/receive > semantics" did not work always i

Re: [OMPI devel] 1.6.1rc1 posted

2012-08-03 Thread Paul Kapinos
Jeff, All, testing our well-known example of the registered memory problem (see http://www.open-mpi.org/community/lists/users/2012/02/18565.php) on freshly-installed 1.6.1rc2, found out that "Fall back to send/receive semantics" did not work always it. However the behaviour has changed: 1.5.

Re: [OMPI devel] 1.6.1rc1 posted

2012-07-30 Thread Jeff Squyres
1.6.1rc1 is a bust because of a compile error. :( It wasn't caught on the build machine because it's a bug in the openib BTL, and the build machine doesn't have OpenFabrics support. 1.6.1rc2 will be posted later today. On Jul 27, 2012, at 10:20 PM, Jeff Squyres wrote: > Finally! It's in th

[OMPI devel] 1.6.1rc1 posted

2012-07-27 Thread Jeff Squyres
Finally! It's in the usual place: http://www.open-mpi.org/software/ompi/v1.6/ Please test, especially with low-registered-memory-available scenarios with Mellanox OpenFabrics devices. Here's a list of changes since 1.6: - A bunch of changes to eliminate hangs on OpenFabrics-based networks