Date: Sat, 16 Mar 2002 18:11:15 -0800 (PST) From: Harkirat Singh <singh@pdx.edu> To: Julian Elischer <julian@elischer.org> Cc: Luigi Rizzo <rizzo@icir.org>, Mike Silbersack <silby@silby.com>, Brian White <bwhite@moab.cs.utah.edu>, <freebsd-net@FreeBSD.ORG> Subject: Re: SACK (and older TCP stack) availability? Message-ID: <Pine.GSO.4.33.0203161736150.4829-100000@gere.odin.pdx.edu> In-Reply-To: <Pine.BSF.4.21.0202211427100.69443-100000@InterJet.elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello, I am sorry due to some reasons I have not been able to check this mail account in the past. I faced problem in validation of Implementation due to lack of efficient tools, I used wireless bed so by default tcpdump (at wi0 interface) was in promiscuous-mode and one packet was dumped appx three times. I used Ostermann's tcptrace and tcpdump and then did a packet by pkt trace, SACK was behaving as desirable. However, I wish if there would have been more efficient tool. I also verified implementation in a wireless testbed of 3 hop network and using Dummynet for creating reorder, loss, delay etc..(many thanks to Rizzo for such a magnificent tool!). SACK showed higher thruput than Reno and Newreno under various test conditions. I was doing an Energy performance of Reno, NewReno and SACK part of my thesis and I used kernel.GENERIC (FreeBSD4.3) for reno and NewReno. Please let me know if I can be of any of help. Regards, Harkirat On Thu, 21 Feb 2002, Julian Elischer wrote: > tehn we need to tell teh author.. > he seems to want then to be used . I'm sure he'd be reponsive to being > told about the problems.. > > > On Thu, 21 Feb 2002, Luigi Rizzo wrote: > > > On Thu, Feb 21, 2002 at 01:54:20AM +0000, Mike Silbersack wrote: > > > > > > > I tracked a thread from Aug 2001 in which Harkirat Singh announced his > > > > SACK implementation. But, this thread seems to have ended prematurely and > > > > grep'ing for SACK in current does not turn up anything interesting. What > > > > is it's status? Was the 4.3 SACK diff stable? > > > > > > I never got time to test it, nor did any other committer I'm aware of. > > > If someone were to do some good testing of the patch and post the results, > > > that would be highly helpful in getting it on the road to being committed. > > > Unadvocated patches ususally don't get committed. > > > > I actually looked at the patches, and by visual inspection, > > they broke the flow for standard TCP connections when SACK > > was disabled. This was also verified with TBIT. > > So even if the SACK implementation was correct > > (which I haven't checked in detail) they are a no-go > > unless someone puts significant work on them. > > > > cheers > > luigi > > ----------------------------------+----------------------------------------- > > Luigi RIZZO, luigi@iet.unipi.it . ICSI (on leave from Univ. di Pisa) > > http://www.iet.unipi.it/~luigi/ . 1947 Center St, Berkeley CA 94704 > > Phone: (510) 666 2927 > > ----------------------------------+----------------------------------------- > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > > with "unsubscribe freebsd-net" in the body of the message > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.GSO.4.33.0203161736150.4829-100000>