The bottom line is that we're talking about carrying an RDMA protocol over an Ethernet wire. It just so happens that this particular proposal uses IB's RDMA protocol. It does so for two reasons; 1. It already exists and has been validated and 2. it is relatively straightforward to implement.
Your example of building an IB router for carrying IB over the WAN is an interesting one. It so happens that I've been working on just such a project; that project has moved away from calling the device an IB router in favor of calling it a WAN gateway. In any case, the analogy is imperfect since the function of the WAN gateway is to connect either two IB subnets, or in the case of the project I've been working on, it is to connect two segments of an IB subnet. In either case, it's clearly about IB. I'll concede the point about FCoE. Nevertheless, I see lots of merit, at least in this case, in using an expression to describe the technology which is crisp and unambiguous. -Paul -----Original Message----- From: Roland Dreier [mailto:[email protected]] Sent: Tuesday, June 16, 2009 2:14 PM To: Paul Grun Cc: 'Jason Gunthorpe'; 'Sean Hefty'; 'ewg'; 'general-list' Subject: Re: [ewg] Re: [ofa-general] [PATCH 0/9] RDMAoE - RDMA over Ethernet > 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. Seems pretty silly to me -- if I build an IB router that carries IB over the WAN on MPLS, is it not InfiniBand because it uses the wrong layer 1? Similarly the worlds seems to cope with talking about FCoE rather than SCSIoE even though FCoE does not share all the OSI layers of fibre channel. _______________________________________________ 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
