If I might chime in here...I've been working to actively squash the expression 'IBoE' or any variation that includes IB in the name. The reason is because the InfiniBand Architecture is defined as a cohesive solution that includes five layers of the OSI stack. Hence using the expression IBoE creates unnecessary confusion by implying that RDMAoE includes elements of IB from the five OSI layers, which is not an accurate description of the proposal. -Paul Grun
-----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Jason Gunthorpe Sent: Tuesday, June 16, 2009 1:56 PM To: Sean Hefty Cc: Roland Dreier; ewg; general-list Subject: [ewg] Re: [ofa-general] [PATCH 0/9] RDMAoE - RDMA over Ethernet On Tue, Jun 16, 2009 at 09:32:25AM -0700, Sean Hefty wrote: > >RDMA over Ethernet (RDMAoE) allows running the IB transport protocol over > >Ethernet, providing IB capabilities for Ethernet fabrics. The packets are > >standard Ethernet frames with an Ethertype, an IB GRH, unmodified IB transport > >headers and payload. HCA RDMAoE ports are no different than regular IB ports > >from the RDMA stack perspective. > > I would refer to this as IBoE, not RDMAoE. I disagree - this is very clearly not IB. Without interop at the SA, CM and other management layers it is effectively a different protocol - that only shares the same wire format for established QPs as IB. Like FCoE the real challenge with something like this is all the surrounding management and interop/interwork with the established base (ie actual IB). As Jim said, IBTA is going to be taking this up, and I'd be really disappointed if the resulting specification was something that could not interwork with IB at all layers. > The RDMA stack should see these ports different than regular IB HCA ports. > There are a lot of differences that should not simply be hidden or incorrectly > assumed: QP0, QoS, multiple paths, routing(?), no SA, etc. .. Or it should actually do all this in some sane IB compatible way .. Jason _______________________________________________ ewg mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg _______________________________________________ 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
