Subject: Re: TCP/SACK support.
To: Christos Zoulas <firstname.lastname@example.org>
From: Jonathan Stone <email@example.com>
Date: 01/06/2005 11:31:27
In message <firstname.lastname@example.org>Christos Zoulas writes
>In article <20050106093004.GB69962@sekhmet.sigusr1.org>,
>Kentaro A. Kurahone <email@example.com> wrote:
>>Someone gently clued me in that this was needed.
>Thanks! We really needed this! Can someone please review it?
I can certainly offer to have a look.
But there are a number of other implementations of SACK for the BSD
TCP stack, some of which have been widely tested, deployed, and not
least, had some review by leading members TCP/IP research community.
[[I dont know how much, but perhaps some of the list members who I
know are closer to those implementations could comment.]]
After considerable thought, I'd strogly prefer to see NetBSD
adopt/import one of the extant versions of of SACK, than to attempt to
roll (and debug) our own in a tight timeframe. I see's considerable
benefit in sharing a common SACK implemetation, and a nontrivial
downside in adding further divergence (I avoid the phrase "gratuitious
divergence") to NetBSD's TCP stack.
I have already spent some time back-porting portions of the FreeBSD-5
TCP SACK code. There's also the DragonFly BSD SACK code. (and of
course the Mathis &c PSU code).
I have also been quietly asking members of the TCP research community
how NetBSD can best proceed with SACK. The impression I get is that
we'd be better going with a variant of an established SACK implementation,
than inventing yet another variant.
Of course, YMMV.