[ofa-general] Re: [ewg] OFED March 24 meeting summary on OFED 1.4 plans

Edward Mascarenhas eddiem at sgi.com
Wed Apr 16 15:27:14 PDT 2008


The SGI Altix ICE cluster system supports 2 InfiniBand fabrics.
http://www.sgi.com/products/servers/altix/ice/
Each compute node has 2 HCAs and each is connected to a separate 
fabric. We recommend that users use one fabric for storage traffic and 
the other for MPI, but there is no reason why both fabrics could not 
be used for MPI. OpenMPI requires setting a separate subnet prefix for 
each fabric to use both fabrics for MPI and OpenSM supports this 
setting of subnet prefix. Other MPIs do not require this.

Edward


on 04/04/2008 08:08 AM Tang, Changqing said the following:
> What I mean "claim to support" is to have more people to test with this config.
> 
> --CQ
> 
>> -----Original Message-----
>> From: Or Gerlitz [mailto:or.gerlitz at gmail.com]
>> Sent: Thursday, April 03, 2008 11:18 PM
>> To: Tang, Changqing
>> Cc: general at lists.openfabrics.org; ewg at lists.openfabrics.org
>> Subject: Re: [ofa-general] Re: [ewg] OFED March 24 meeting
>> summary on OFED 1.4 plans
>>
>> On Thu, Apr 3, 2008 at 5:40 PM, Tang, Changqing
>> <changquing.tang at hp.com> wrote:
>>
>>>  The problem is, from MPI side, (and by default), we don't
>> know which
>>> port is on which  fabric, since the subnet prefix is the
>> same. We rely
>>> on system admin to config two  different subnet prefixes
>> for HP-MPI to work.
>>>  No vendor has claimed to support this.
>> CQ, not supporting a different subnet prefix per IB subnet is
>> against IB nature, I don't think there should be any problem
>> to configure a different prefix at each open SM instance and
>> the Linux host stack would work perfectly under this config.
>> If you are a ware to any problem in the opensm and/or the
>> host stack please let the community know and the maintainers
>> will fix it.
>>
>> Or.
>>
> _______________________________________________
> ewg mailing list
> ewg at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg



More information about the ewg mailing list