Hey Yevgeny,

> Did ftree fail to digest the topology? 

For the current OpenSM, on atleast one cluster, yes.  I've been meaning
to look into it.  Just haven't gotten to it yet.

There is some legacy too. Earlier ftree's weren't able to handle a
number of corner cases, so updn was used instead (and now is still
used).

> Or perhaps that you need LMC>0, which ftree doesn't support?

This too.

Al

On Wed, 2008-05-28 at 09:24 +0300, Yevgeny Kliteynik wrote:
> Ira,
> 
> Ira Weiny wrote:
> > 
> > We run with the up/down algorithm, ftree has not panned out for us yet.
> 
> Can you elaborate on that?
> Did ftree fail to digest the topology? Or did it do a lousy job configuring
> the subnet? Or perhaps the you need LMC>0, which ftree doesn't support?
> 
> -- Yevgeny
> 
> _______________________________________________
> general mailing list
> [email protected]
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
> 
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
-- 
Albert Chu
[EMAIL PROTECTED]
925-422-5311
Computer Scientist
High Performance Systems Division
Lawrence Livermore National Laboratory

_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general

To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

Reply via email to