[ofa-general] Limited number of multicasts groups that can be joined?

[Hal Rosenstock]

On 7/19/07, Andrew Friedley <afriedle at open-mpi.org> wrote:
>
>
>
> Hal Rosenstock wrote:
> > Thanks. I can only comment on the OpenSM configuration and in general on
> > SMs
> > so I'm still not sure what limits you are hitting; it may be multiple
> but
> > not sure. Some seemed to be end node (HCA) related based on a previous
> > email.
>
> Thanks for you help.  Yes I'm thinking the same thing, though what I'm
> seeing seemingly contradicts the limits that I'm told are in place (and
> have now been changed post-v1.2).
>
> >> Is the algorithm for recalculating the tree documented at all?  Or,
> >> where is the code for it (assuming I have access)?  I feel like I'm
> >> missing something here that explains why it's so costly.
> >
> >
> > I'm afraid it is just the code AFAIK :-(
>
> OK, do you know where it is in the OpenSM code base?


Start with osm_sa_mcmember_record.c and work towards:
 osm_mcast_mgr.c  osm_mcm_info.c  osm_mtree.c
 osm_mcast_tbl.c  osm_mcm_port.c  osm_multicast.c

-- Hal

Andrew
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20070719/8d29f309/attachment.html>