From owner-freebsd-net Fri Aug 4 14: 8:14 2000 Delivered-To: freebsd-net@freebsd.org Received: from bmah-freebsd-0.cisco.com (bmah-freebsd-0.cisco.com [171.70.84.42]) by hub.freebsd.org (Postfix) with ESMTP id E224537B75B for ; Fri, 4 Aug 2000 14:08:05 -0700 (PDT) (envelope-from bmah@cisco.com) Received: (from bmah@localhost) by bmah-freebsd-0.cisco.com (8.11.0/8.11.0) id e74L7Tu13995; Fri, 4 Aug 2000 14:07:29 -0700 (PDT) (envelope-from bmah) Message-Id: <200008042107.e74L7Tu13995@bmah-freebsd-0.cisco.com> X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: Tarik Alj Cc: archie@whistle.com, freebsd-net@FreeBSD.ORG Subject: Re: cvs commit: src/sys/netinet tcp_output.c In-Reply-To: <200008042014.QAA05452@cholla.INRS-Telecom.UQuebec.CA> References: <200008042014.QAA05452@cholla.INRS-Telecom.UQuebec.CA> Comments: In-reply-to Tarik Alj message dated "Fri, 04 Aug 2000 16:14:31 -0400." From: bmah@cisco.com (Bruce A. Mah) Reply-To: bmah@cisco.com X-Face: g~c`.{#4q0"(V*b#g[i~rXgm*w;:nMfz%_RZLma)UgGN&=j`5vXoU^@n5v4:OO)c["!w)nD/!!~e4Sj7LiT'6*wZ83454H""lb{CC%T37O!!'S$S&D}sem7I[A 2V%N&+ X-Image-Url: http://www.employees.org/~bmah/Images/bmah-cisco-small.gif X-Url: http://www.employees.org/~bmah/ Mime-Version: 1.0 Content-Type: multipart/signed; boundary="==_Exmh_2103521206P"; micalg=pgp-sha1; protocol="application/pgp-signature" Content-Transfer-Encoding: 7bit Date: Fri, 04 Aug 2000 14:07:29 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org --==_Exmh_2103521206P Content-Type: text/plain; charset=us-ascii If memory serves me right, Tarik Alj wrote: > > > > > >Hmm.. Can you explain what "RED" is and how the change breaks it? > > RED stands for Random Early Detection, you can find more info about it (at le > ast > a bunch of pointers) from: [http://www.aciri.org/floyd/red.html]. To answer the other half of the question, I think the problem is that if the queue is full, the committed patch calls to drop the latest outbound packet. That's the definition of "drop tail". A RED queue would drop a random packet from the full queue. The idea, in a really handwavy way, is that flows with many packets enqueued are more susceptable to having something dropped. If you can get those flows to back off (e.g. TCP congestion control), that'll hopefully have the most effect of alleviating the congestion in the queue. Garrett is probably right in a purist sense, but if no one's going to have RED in the stack (e.g. ALTQ) it seems this patch fixes at least *something*. Bruce. PS. Hi Archie! --==_Exmh_2103521206P Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 5.0i for non-commercial use MessageID: elU4ojSwEQblVPaewQ6LsY8kNLlDkOhT iQA/AwUBOYswkdjKMXFboFLDEQJwwgCfY16VqxdQt35OLMG62kPt4XISFb4An2wM 0xfuxIfYOS6lYml3KlYbOJji =/3rr -----END PGP SIGNATURE----- --==_Exmh_2103521206P-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message