|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: Enough on TCP vs. FCTP, pleaseDavid: Let me take a rain-check on this issue... I want to go poke around in a couple of different TCP implemenations and contrast that with Matt's draft.... After doing that I will get back to this thead on my opinion :) I think a study of: A) How current implemenations apply the urgent pointer, B) How this contrasts with Matt's draft, AND C) How all of (A) and (B) contrast with RFC793 must be considered... Now only after a careful assessment of all three will I chime in... My schedule is a bit tight until the end of the week when I will have more time to devote to this subject.. Thats all for now... and by the way Julian and Matt: 1) I have only commented to this WG on SCTP when I deemed it appropriate. Mainly because no one wants to go out and read the RFC/draft on it (it will be published later this week) to figure out what additional features it can apply. 2) I will continue to do so as I see fit.. Not that I have put as many comments as you seem to think I have ... Regards R Black_David@emc.com wrote: > > > End schemes to fix TCP; address the issues of TCP and SCTP. Do not stand > on > > desires for TCP over SCTP or that a pact was made to promote TCP. SCTP > > performance will be superior to TCP using either hardware or software. > SCTP > > is easier to accelerate in hardware than TCP and any resulting SCSI > standard > > will change substantially as a result of the features of SCTP. Even if > > there was a generic RDMA option, SCSI would look more like Firewire as a > > result. A debased version of RDMA has already been offered that avoids > > stepping on accepted options. > > I think this sort of debate really needs to stop. Matt's draft is a > reasonable > attempt to apply an existing TCP feature to address a shortcoming in use > of TCP for iSCSI. It is not perfect, as is the case for most engineering > tradeoffs and compromises. SCTP does have a number of improvements > over TCP (e.g., the whole session discussion is moot for SCTP), but TCP > has significantly more deployment experience and a great deal of work in > progress on hardware acceleration. > > I am restating the WG consensus that iSCSI needs to anticipate both > SCTP and TCP, with the assumption that TCP deployments will come > first. Given this, Matt's draft is within scope for the WG. If anyone > other > than Doug Otis disagrees, please say so and say why on the mailing list. > > --David > > --------------------------------------------------- > David L. Black, Senior Technologist > EMC Corporation, 42 South St., Hopkinton, MA 01748 > +1 (508) 435-1000 x75140 FAX: +1 (508) 497-8500 > black_david@emc.com Mobile: +1 (978) 394-7754 > --------------------------------------------------- -- Randall R. Stewart randall@stewart.chicago.il.us or rrs@cisco.com 815-342-5222 (cell) 815-477-2127 (work)
Home Last updated: Tue Sep 04 01:06:55 2001 6315 messages in chronological order |