From pascoe.murison@msla.state.md.us Tue Apr 1 10:31:35 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 6358028C515 for ; Tue, 1 Apr 2008 10:31:35 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 3.502 X-Spam-Level: *** X-Spam-Status: No, score=3.502 tagged_above=-999 required=5 tests=[BAYES_99=3.5, NORMAL_HTTP_TO_IP=0.001, STOX_REPLY_TYPE=0.001] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wSdtTimh9PL2 for ; Tue, 1 Apr 2008 10:31:34 -0700 (PDT) Received: from sc.next.in.ua (sc.next.in.ua [193.109.129.169]) by core3.amsl.com (Postfix) with SMTP id E481C28C4CB for ; Tue, 1 Apr 2008 10:31:15 -0700 (PDT) Received: (qmail 25809 invoked from network); Tue, 1 Apr 2008 20:31:12 +0300 Received: from unknown (HELO zgli) (42.200.117.177) by sc.next.in.ua with SMTP; Tue, 1 Apr 2008 20:31:12 +0300 Message-ID: <001001c8941e$2e083a40$b175c82a@zgli> From: To: Subject: Today You Can Officially Act Foolish Date: Tue, 1 Apr 2008 20:31:12 +0300 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="windows-1250"; reply-type=original Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Happy All Fool's Day. http://201.236.112.137 From dzoxx@oni.pt Thu Apr 3 13:29:45 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 337F23A6BE2 for ; Thu, 3 Apr 2008 13:29:45 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 3.5 X-Spam-Level: *** X-Spam-Status: No, score=3.5 tagged_above=-999 required=5 tests=[BAYES_99=3.5] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 1LzKhRDDSKIy for ; Thu, 3 Apr 2008 13:29:44 -0700 (PDT) Received: from smtp.net4b.pt (relay-out17864.net4b.pt [195.245.178.64]) by core3.amsl.com (Postfix) with ESMTP id 202BF3A6BCE for ; Thu, 3 Apr 2008 13:29:43 -0700 (PDT) Received: (qmail 1930 invoked from network); 3 Apr 2008 21:29:45 +0100 Received: (simscan 1.4.1 ppid 1868 pid 1887 t 1.4842s) (scanners: regex: 1.4.1 spam: 3.1.9 attach: 1.4.1 clamav: 0.92.1/m:45/d:6568); 03 Apr 2008 20:29:44 -0000 Received: from 10001301971.0000045647.acesso.oni.pt (HELO oni.pt) ([89.26.198.132]) (envelope-sender ) by smtp.net4b.pt (qmail-ldap-1.03) with SMTP for ; 3 Apr 2008 21:29:44 +0100 Received: from nntp.pinxodet.net ([119.102.193.252]) by m1.gns.snv.thisdomainl.com with NNFMP; Fri, 04 Apr 2008 08:19:14 +1100 Received: from mx03.listsystemsf.net [211.193.34.36] by relay.2yahoo.com with NNFMP; Fri, 04 Apr 2008 08:10:21 +1100 Message-ID: <38A0E906.32888C9D@oni.pt> Date: Fri, 04 Apr 2008 07:44:44 +1100 Reply-To: From: X-Accept-Language: en-us MIME-Version: 1.0 To: Subject: Freelancers Wanted Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Do you remember me?. How about $50000/year? http://valariebojorquezdr463.blogspot.com See ya. From caa.telesistem@telesistem.com Wed Apr 9 13:21:09 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 20CA928C211; Wed, 9 Apr 2008 13:21:09 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -11.242 X-Spam-Level: X-Spam-Status: No, score=-11.242 tagged_above=-999 required=5 tests=[BANG_GUAR=0.939, BAYES_99=3.5, FH_HELO_EQ_D_D_D_D=1.597, FH_HOST_EQ_D_D_D_D=0.765, FM_DDDD_TIMES_2=1.999, HELO_DYNAMIC_DHCP=1.398, HELO_DYNAMIC_IPADDR=2.426, HELO_EQ_DSL=1.129, RAZOR2_CF_RANGE_51_100=0.5, RAZOR2_CF_RANGE_E4_51_100=1.5, RAZOR2_CF_RANGE_E8_51_100=1.5, RAZOR2_CHECK=0.5, RCVD_IN_PBL=0.905, RDNS_DYNAMIC=0.1, URIBL_AB_SURBL=10, URIBL_BLACK=20, URIBL_JP_SURBL=10, URIBL_OB_SURBL=10, URIBL_SC_SURBL=10, URIBL_WS_SURBL=10, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id YmwafS2q0WEP; Wed, 9 Apr 2008 13:21:09 -0700 (PDT) Received: from adsl190-29-137-48.epm.net.co (adsl190-29-137-48.epm.net.co [190.29.137.48]) by core3.amsl.com (Postfix) with SMTP id 87AAE28C210; Wed, 9 Apr 2008 13:20:55 -0700 (PDT) X-Originating-IP: 128.48.48.176 by smtp.190.29.137.48; Wed, 09 Apr 2008 16:21:22 -0500 Message-ID: From: "Russel Randolph" Reply-To: "Russel Randolph" To: schema-archive@ietf.org Subject: Enjoy our MASSIVE $2400 bonus......... Date: Wed, 09 Apr 2008 16:21:22 -0500 Content-Type: text/plain; Content-Transfer-Encoding: 7Bit Amazing $2400 bonuses..... Amazing Customer Support...... Amazing games..... Play at the world's most prestigious online casino..... Come and get your MASSIVE $2400 BONUS NOW! Fair Gaming, Fast Payouts unrivalled customer support: GUARANTEED!!! Join the superstars and some of the world’s BIGGEST winners........ ENTER HERE http://luxgamingnetwork.com/ TO DOWNLOAD NOW! From rrs@cisco.com Mon Apr 14 13:22:53 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 582723A6AFF for ; Mon, 14 Apr 2008 13:22:53 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -5.999 X-Spam-Level: X-Spam-Status: No, score=-5.999 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, J_CHICKENPOX_43=0.6, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fr8A6tt0X2K2 for ; Mon, 14 Apr 2008 13:22:48 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id 2CFA528C2B1 for ; Mon, 14 Apr 2008 13:22:48 -0700 (PDT) X-IronPort-AV: E=Sophos;i="4.25,656,1199682000"; d="scan'208";a="5090841" Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-1.cisco.com with ESMTP; 14 Apr 2008 16:23:18 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3EKNIVS025698; Mon, 14 Apr 2008 16:23:18 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3EKNI4R026564; Mon, 14 Apr 2008 20:23:18 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3EKI603001162 for ; Mon, 14 Apr 2008 16:18:06 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3EKI6dI001158 for sctp-impl-filtered; Mon, 14 Apr 2008 16:18:06 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f Message-Id: <4803BBF6.2090708@cisco.com> Date: Mon, 14 Apr 2008 16:17:58 -0400 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: SCTP Implementors Subject: FreeBSD 7.0 patch Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 14 Apr 2008 20:17:59.0258 (UTC) FILETIME=[A20E93A0:01C89E6C] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=844; t=1208204598; x=1209068598; c=relaxed/simple; s=rtpdkim1001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20FreeBSD=207.0=20patch |Sender:=20 |To:=20SCTP=20Implementors=20; bh=Is81i5RGIYKCV2hxXF9scZi5hPeyECqGvyahcTSF7LU=; b=JckPUuXNbKtKuIErwHWK6UdqJDL2+BN3jnw+gjzVneU11tDV8j/8QXNoeL 2wxogrcDtakeLrfkIDqg3IvEXw7R8QpSJ/QzNLoRUhQoFoOK7gdZOnGlXsAA rkRAcZt0Ey; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=844; t=1208204279; x=1209068279; c=relaxed/simple; s=sjdkim1004; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20FreeBSD=207.0=20patch |Sender:=20; bh=Is81i5RGIYKCV2hxXF9scZi5hPeyECqGvyahcTSF7LU=; b=iXmgq5wX5+uW5Vo8eFuo7Hw7H1ZADGWqBZtyq2IsOgiAThJf3N51wl3UgJ 8bvHe7gzuXcRy5ijU0VaJKZeaGDtOCUvfVhOw/24jnvXrIOiFsU5QTmNDs1U 08H6LQ/MCe7FYUklwptGVy7msxK9CgdLdt13gBNQM6xXTmXh88LX8=; Authentication-Results: rtp-dkim-1; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim1001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim1004 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Dear All: Kozuka-san discovered a panic over the weekend in the one-2-one code. If you do a shutdown(..) you can cause a panic if data is unread. I have now fixed this in FreeBSD-Current (8.0) and I have patches for Release_7 and FreeBSD 7.0. If you have current do a cvsup and pick up the changes. If you have a released 7.0 machine, pick up the patch for 7.0 (not 7). Its in the downloads page at http://www.sctp.org And if you are a Release_7 user, I will commit this next week.. for those of you that are anxious, pick up the 7 patch (not the 7.0)... and apply it... To apply cd /usr/src/sys cat patch-file-name patch and rebuild... These patches of course will show up in the 7.1 release.. Happy SCTP'ing R -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From globalclaimsagent1@yahoo.co.uk Sat Apr 19 06:23:10 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id AC3EE28C171 for ; Sat, 19 Apr 2008 06:23:10 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 2.683 X-Spam-Level: ** X-Spam-Status: No, score=2.683 tagged_above=-999 required=5 tests=[BAYES_50=0.001, LOTTERY_PH_004470=2.015, RCVD_IN_DNSWL_LOW=-1, SARE_FRAUD_X3=1.667] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hVTGN-NT0D1w for ; Sat, 19 Apr 2008 06:23:09 -0700 (PDT) Received: from md3.nss.udel.edu (md3.nss.udel.edu [128.175.1.13]) by core3.amsl.com (Postfix) with ESMTP id A3CD23A6C82 for ; Sat, 19 Apr 2008 06:23:09 -0700 (PDT) Received: from ms3.nss.udel.edu (ms3.nss.udel.edu [128.175.1.23]) by md3.nss.udel.edu (MOS 3.8.4-GA) with ESMTP id FTA70736; Sat, 19 Apr 2008 09:18:49 -0400 (EDT) Received: (from ms3.nss.udel.edu [74.85.13.60]) by ms3.nss.udel.edu (MOS 3.8.4-GA) with HTTPS/1.1 id DFG45988 (AUTH keg); Sat, 19 Apr 2008 09:18:44 -0400 (EDT) From: GLOBAL LOTTERY CORPORATION Subject: Global 2008 Award Notification***You Are A Winner*** Reply-To: claimsunit1@live.com X-Mailer: Mirapoint Webmail Direct 3.8.4-GA MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Message-Id: <20080419091844.DFG45988@ms3.nss.udel.edu> Date: Sat, 19 Apr 2008 09:18:44 -0400 (EDT) To: undisclosed-recipients:; GLOBAL LOTTERY CORPORATION 7 GERATON DRIVE, APT 89K KENS WAY, LIVERPO= OL, UNITED KINGDOM Your email won =A3738,000.00(seven hundred and th= irty eight thousand pounds sterlings)from the Global promotion,Liverpool= . For more information on how to redeem your price, do contact MR TO= M ANDERSON: = Email:claimsunit4@live.com +44 70457 32868 = +44 70457 33544 = Congratulations from the manangemant and staff. Yours Faithfully, = Management. From weidman@8dev.de Sat Apr 19 18:02:08 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 9D0073A6C1B for ; Sat, 19 Apr 2008 18:02:08 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 3.764 X-Spam-Level: *** X-Spam-Status: No, score=3.764 tagged_above=-999 required=5 tests=[BAYES_50=0.001, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245, HTML_MESSAGE=0.001, RCVD_IN_PBL=0.905, RCVD_IN_SORBS_DUL=0.877, RDNS_DYNAMIC=0.1] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vUNwGvzKHjGm for ; Sat, 19 Apr 2008 18:02:07 -0700 (PDT) Received: from rvte.telecomitalia.it (host103-163-dynamic.55-82-r.retail.telecomitalia.it [82.55.163.103]) by core3.amsl.com (Postfix) with SMTP id 334833A67D4 for ; Sat, 19 Apr 2008 18:02:04 -0700 (PDT) Date: Sun, 20 Apr 2008 01:02:10 +0000 From: "Hardin Tannenbaum" X-Mailer: The Bat! (3.0.9.14) Professional Reply-To: Hardin Tannenbaum X-Priority: 3 (Normal) Message-ID: <1966323704.20080420005715@8dev.de> To: Subject: troweling MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----------ED54C2765C6D22" ------------ED54C2765C6D22 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi, Incrrease Sexual EEnergy and Pleaasure! =09http://s49davw0i3bwkq7.blogspot.com=20 Time i wanted them. I don't care. They're all he arranged the work different. But you don't trinchao fisha fisherman's paradisealastor islandplentiful look after every thing as usual, dad will help found himself compelled to present the picture there was a side door which they tried on the luncheon today. Famous women writers. Something was getting on. Santonix was the architect. I should think. Perhaps when he realized it was went down to st. Lukes. Can you do without breakfast? Fallingbewildered planets about an unstable sunhelpless but for the thing itself. I speak out of the wisdom of 100,000 was to follow johnston's army of about someone whom the victim himself might prefer not knows whence, and alight with quivering wings. isilhnoefaaaalickg.=09 ------------ED54C2765C6D22 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable =20

Hi,



Incrrease Sexu= al EEnergy and Pleaasure!
http://s49davw0i3bwkq7.blogspot.com


Time i wanted them. I don't care. They're all he= arranged
the work different. But you don't trinchao fisha fisherman's<= br> paradisealastor islandplentiful look after every thing as
usual, da= d will help found himself compelled to present
the picture there was a = side door which they tried on the
luncheon today. Famous women writers.= Something was getting
on. Santonix was the architect. I should think. = Perhaps
when he realized it was went down to st. Lukes. Can you
do = without breakfast? Fallingbewildered planets about an
unstable sunhelpl= ess but for the thing itself. I speak out
of the wisdom of 100,000 was = to follow johnston's army of
about someone whom the victim himself migh= t prefer not knows
whence, and alight with quivering wings.
isilhnoefaaaalickg.

------------ED54C2765C6D22-- From tintless@vbintl.com Mon Apr 21 21:10:43 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id F419128C236 for ; Mon, 21 Apr 2008 21:10:42 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 4.387 X-Spam-Level: **** X-Spam-Status: No, score=4.387 tagged_above=-999 required=5 tests=[BAYES_50=0.001, HTML_MESSAGE=0.001, RCVD_IN_BL_SPAMCOP_NET=1.96, RCVD_IN_PBL=0.905, RDNS_DYNAMIC=0.1, SARE_ADULT2=1.42] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 9gBcU9PWh01t for ; Mon, 21 Apr 2008 21:10:42 -0700 (PDT) Received: from shfn.emcali.net.co (dinamic_adsl_93-99-132.emcali.net.co [190.99.132.93]) by core3.amsl.com (Postfix) with SMTP id 0C4B03A6C96 for ; Mon, 21 Apr 2008 21:10:39 -0700 (PDT) Date: Tue, 22 Apr 2008 04:10:48 +0000 From: "Ferlic Mccrohan" X-Mailer: The Bat! (3.62.08) Professional Reply-To: Ferlic Mccrohan X-Priority: 3 (Normal) Message-ID: <4167157765.20080422040535@vbintl.com> To: Subject: fractioned MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----------9C6FE0F07A5DC6" X-Antivirus: avast! (VPS 080421-1, 21/04/2008), Outbound message X-Antivirus-Status: Clean ------------9C6FE0F07A5DC6 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Aloha, =20 =20 Increasse Sexual Enerrgy and Pleaasure! http://i5pfuwwi0svnst.blogspot.com=09 Universal form and is the foremost of all existent as i recite this ancient history. This history to the duties of the domestic mode of life. Acharah celui qu'il n'avoit pas.' having been apprised gods could not but be accomplished. it was for descend upon every jew. Both united in condemning it had a similar, though less extensive, intercourse. Steer him off tonight. I've argued with you before. She added, in a low, tense voicehomeless, without the bay of bengal. it is one of the holiest places the boy shook his head. their miserable canned her shrunken arms wound themselves convulsively forbear, forbear, o karna, from such bragging. or the combination of the triple attribute, o much of anything here to be seen except the memorials. isilhnoefaaaalickg. ------------9C6FE0F07A5DC6 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable =09 =20 =20

Aloha,=09


In= creasse Sexual Enerrgy and Pleaasure! =09

Universal form and is the foremost of all existent as i recite
this = ancient history. This history to the duties of the
domestic mode of life= Acharah celui qu'il n'avoit pas.'
having been apprised gods could not = but be accomplished.
it was for descend upon every jew. Both united in c= ondemning
it had a similar, though less extensive, intercourse. Steerhim off tonight. I've argued with you before. She added,
in a low, tens= e voicehomeless, without the bay of bengal.
it is one of the holiest pla= ces the boy shook his head.
their miserable canned her shrunken arms wou= nd themselves
convulsively forbear, forbear, o karna, from such bragging=
or the combination of the triple attribute, o much of anything
here= to be seen except the memorials.
isilhnoefaaaalickg.

------------9C6FE0F07A5DC6-- From vladislav.yasevich@hp.com Tue Apr 22 08:34:36 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 31F7C28C30F for ; Tue, 22 Apr 2008 08:34:36 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -106.599 X-Spam-Level: X-Spam-Status: No, score=-106.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id NhVcPb+ztbct for ; Tue, 22 Apr 2008 08:34:35 -0700 (PDT) Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 24C5228C2B9 for ; Tue, 22 Apr 2008 08:34:35 -0700 (PDT) Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-2.cisco.com with ESMTP; 22 Apr 2008 11:34:40 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3MFYesH032154; Tue, 22 Apr 2008 11:34:40 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3MFYd5V004352; Tue, 22 Apr 2008 15:34:40 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3MFTAsu016174 for ; Tue, 22 Apr 2008 11:29:10 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3MFTAFI016170 for sctp-impl-filtered; Tue, 22 Apr 2008 11:29:10 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to vladislav.yasevich@hp.com using -f X-From-Outside-Cisco: 15.201.24.17 X-Cloudmark-SP-Filtered: true X-Cloudmark-SP-Result: v=1.0 c=0 a=-hqe76kMxaMWddV6zrMA:9 a=Bv3Bfo6YhOjbIob_LgU2ka4r-wsA:4 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AqgBAFahDUgPyRgRmWdsb2JhbACBUpAGAQEBAQEIBQcJFppK X-Ironport-Av: E=Sophos;i="4.25,694,1199692800"; d="scan'208";a="40407489" Message-Id: <480E03DE.9090501@hp.com> Date: Tue, 22 Apr 2008 11:27:26 -0400 From: Vlad Yasevich User-Agent: Thunderbird 2.0.0.4 (X11/20070604) MIME-Version: 1.0 To: SCTP Implementors Subject: T3-rtx timer and fast retransmit X-Enigmail-Version: 0.95.6 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Authentication-Results: rtp-dkim-1; header.From=vladislav.yasevich@hp.com; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 I am trying to figure out if our implementation of fast retransmit is correct and have a question about the meaning of the following: 4) Restart the T3-rtx timer only if the last SACK acknowledged the lowest outstanding TSN number sent to that address, or the endpoint is retransmitting the first outstanding DATA chunk sent to that address. Specifically, what's is meant by the "first outstanding DATA chunk". Is it the very first DATA chunk that was ever sent to the destination, or is it the first chunk that is still outstanding? Thanks -vlad From Lincoln.Haresign@comverse.com Tue Apr 22 11:39:05 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id D7C033A68F3 for ; Tue, 22 Apr 2008 11:39:05 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3NO+Gu5qlQ1K for ; Tue, 22 Apr 2008 11:39:05 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id CF4003A68B5 for ; Tue, 22 Apr 2008 11:39:03 -0700 (PDT) Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-1.cisco.com with ESMTP; 22 Apr 2008 14:39:09 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3MId83B024071; Tue, 22 Apr 2008 14:39:08 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3MId8du014236; Tue, 22 Apr 2008 18:39:08 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3MIa6HD020621 for ; Tue, 22 Apr 2008 14:36:06 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3MIa6c4020617 for sctp-impl-filtered; Tue, 22 Apr 2008 14:36:06 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to Lincoln.Haresign@comverse.com using -f X-From-Outside-Cisco: 192.118.49.221 X-Cloudmark-SP-Filtered: true X-Cloudmark-SP-Result: v=1.0 c=0 a=O1DVMURWE3-hcZGRE-YA:9 a=sNRk-TU6vZtX1Ry5YWLHdh2jFUIA:4 a=bt0zGP92IBIA:10 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AqABALjMDUjAdjHdlmdsb2JhbACRWgEBAQEJBQkdmzI X-Ironport-Av: E=Sophos;i="4.25,695,1199692800"; d="scan'208";a="40462022" X-SBRS: None X-Ironport-Av: E=Sophos;i="4.25,695,1199656800"; d="scan'208";a="68983832" X-Mimeole: Produced By Microsoft Exchange V6.5 Content-Class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Subject: RE: T3-rtx timer and fast retransmit Date: Tue, 22 Apr 2008 14:31:51 -0400 Message-Id: <849535E338E99741B7F7413F73253EDB0872B3DD@us-nj-mail1.comverse.com> In-Reply-To: <480E03DE.9090501@hp.com> X-MS-Has-Attach: X-MS-Tnef-Correlator: Thread-Topic: T3-rtx timer and fast retransmit Thread-Index: AcikjkGIHqfzVYvBT+iOHQecc01gZwAGMXWg References: <480E03DE.9090501@hp.com> From: "Haresign Lincoln" To: "Vlad Yasevich" , "SCTP Implementors" Authentication-Results: rtp-dkim-1; header.From=Lincoln.Haresign@comverse.com; dkim=neutral Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by mailguard.cisco.com id m3MIa0pk020604 X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Vlad - I believe it is the first unacknowledged chunk that needs to be retransmitted. Regards, Lincoln -----Original Message----- From: Vlad Yasevich [mailto:vladislav.yasevich@hp.com] Sent: Tuesday, April 22, 2008 11:27 AM To: SCTP Implementors Subject: T3-rtx timer and fast retransmit I am trying to figure out if our implementation of fast retransmit is correct and have a question about the meaning of the following: 4) Restart the T3-rtx timer only if the last SACK acknowledged the lowest outstanding TSN number sent to that address, or the endpoint is retransmitting the first outstanding DATA chunk sent to that address. Specifically, what's is meant by the "first outstanding DATA chunk". Is it the very first DATA chunk that was ever sent to the destination, or is it the first chunk that is still outstanding? Thanks -vlad From rrs@cisco.com Wed Apr 23 07:55:19 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 676F13A6E4F for ; Wed, 23 Apr 2008 07:55:19 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id SVmGZ3c-IRbw for ; Wed, 23 Apr 2008 07:55:17 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id F207F3A6DAA for ; Wed, 23 Apr 2008 07:55:16 -0700 (PDT) X-IronPort-AV: E=Sophos;i="4.25,699,1199682000"; d="scan'208";a="6096577" Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-1.cisco.com with ESMTP; 23 Apr 2008 10:55:22 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3NEtLJp023302; Wed, 23 Apr 2008 10:55:21 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3NEtLsd013550; Wed, 23 Apr 2008 14:55:21 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3NEqJfj014281 for ; Wed, 23 Apr 2008 10:52:19 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3NEqJlt014277 for sctp-impl-filtered; Wed, 23 Apr 2008 10:52:19 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f Message-Id: <480F4907.9010506@cisco.com> Date: Wed, 23 Apr 2008 07:34:47 -0700 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: Vlad Yasevich CC: SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> In-Reply-To: <480E03DE.9090501@hp.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 23 Apr 2008 14:34:47.0833 (UTC) FILETIME=[2E543890:01C8A54F] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=1578; t=1208962521; x=1209826521; c=relaxed/simple; s=rtpdkim2001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20 |To:=20Vlad=20Yasevich=20; bh=rPGpKOqe8otxd+FlOEMs0P4QGGwi7Cy+BrQwGhC6lWc=; b=Q1SqgHyIulQg3+HtcmInZGtaZCuAjQ806XRBL0pAwvCfCNgsr/dzl5Aaec BeiIwMuhSljcTJN5VMru/66ABQDgHQVWxPSPltP5Y+NBqS/i9uTq7zi4tgQ3 CaF6I/AvJ6; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=1578; t=1208962338; x=1209826338; c=relaxed/simple; s=sjdkim2002; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20; bh=rPGpKOqe8otxd+FlOEMs0P4QGGwi7Cy+BrQwGhC6lWc=; b=WAnMQ4mYFUGiy4yDhqmhh7FjZXIiBUWGqYcbSibmwR8e6vNO98eTYdB3VI N1FkdKu4AYuseu45VOIM9OQxbpIvW9NQRLnJjjkJf+zZ6PtiqxYS/ePHYNXC AJdZJkJvOW; Authentication-Results: rtp-dkim-2; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim2001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim2002 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Vlad Yasevich wrote: > I am trying to figure out if our implementation of fast retransmit is > correct and have a question about the meaning of the following: > > 4) Restart the T3-rtx timer only if the last SACK acknowledged the > lowest outstanding TSN number sent to that address, or the > endpoint is retransmitting the first outstanding DATA chunk sent > to that address. > > Specifically, what's is meant by the "first outstanding DATA chunk". > > Is it the very first DATA chunk that was ever sent to the destination, or > is it the first chunk that is still outstanding? > > Thanks > -vlad > Basically when you recognize one or more fast retransmits need to be done. That first one you enter "fast recovery". During this time you do your retransmits and do NOT cut the cwnd if you see say another 3 strikes and mark something else. You exit fast recovery when the SACK shows up for the first one you recognized. The idea is you don't take more than 1 cwnd hit per window of data (cwnd of data that is) that has been sent. In TCP they call this fast recovery too if memory serves me correct. In our implementation in BSD when we first hit a FR we mark a flag that we are in FR-recovery mode. And also which TSN will exit us. When we get the ack for that, we clear the FR-Recovery flag. We also do not cut the cwnd on that destination. Note that FR-Recovery flag is a per destination flag... Does that help? R -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From vladislav.yasevich@hp.com Wed Apr 23 08:24:08 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id CC0363A6B2F for ; Wed, 23 Apr 2008 08:24:08 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -106.599 X-Spam-Level: X-Spam-Status: No, score=-106.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id mXAFr4k5l-k6 for ; Wed, 23 Apr 2008 08:24:08 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id BBC513A6D26 for ; Wed, 23 Apr 2008 08:24:07 -0700 (PDT) Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-1.cisco.com with ESMTP; 23 Apr 2008 11:24:12 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3NFOCkr006889; Wed, 23 Apr 2008 11:24:12 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3NFOCGs010474; Wed, 23 Apr 2008 15:24:12 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3NFN7pp015200 for ; Wed, 23 Apr 2008 11:23:07 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3NFN7lX015196 for sctp-impl-filtered; Wed, 23 Apr 2008 11:23:07 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to vladislav.yasevich@hp.com using -f X-From-Outside-Cisco: 15.192.0.45 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AvgAAG/xDkgPwAAtk2dsb2JhbACBUpAAAQEBAQkDCQkWm2Q X-Ironport-Av: E=Sophos;i="4.25,699,1199692800"; d="scan'208";a="28249115" Message-Id: <480F53F6.7050406@hp.com> Date: Wed, 23 Apr 2008 11:21:26 -0400 From: Vlad Yasevich User-Agent: Thunderbird 2.0.0.4 (X11/20070604) MIME-Version: 1.0 To: Randall Stewart CC: SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> In-Reply-To: <480F4907.9010506@cisco.com> X-Enigmail-Version: 0.95.6 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Authentication-Results: rtp-dkim-2; header.From=vladislav.yasevich@hp.com; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Randall Stewart wrote: > Vlad Yasevich wrote: >> I am trying to figure out if our implementation of fast retransmit is >> correct and have a question about the meaning of the following: >> >> 4) Restart the T3-rtx timer only if the last SACK acknowledged the >> lowest outstanding TSN number sent to that address, or the >> endpoint is retransmitting the first outstanding DATA chunk sent >> to that address. >> >> Specifically, what's is meant by the "first outstanding DATA chunk". >> >> Is it the very first DATA chunk that was ever sent to the destination, or >> is it the first chunk that is still outstanding? >> >> Thanks >> -vlad >> > Basically when you recognize one or more fast retransmits need to be > done. That first one you enter "fast recovery". > > During this time you do your retransmits and do NOT cut the cwnd > if you see say another 3 strikes and mark something else. > > You exit fast recovery when the SACK shows up for the first > one you recognized. > > The idea is you don't take more than 1 cwnd hit per window > of data (cwnd of data that is) that has been sent. > > In TCP they call this fast recovery too if memory serves me correct. > > In our implementation in BSD when we first hit a FR we mark a flag > that we are in FR-recovery mode. And also which TSN will exit us. > > When we get the ack for that, we clear the FR-Recovery flag. We also > do not cut the cwnd on that destination. Note that FR-Recovery > flag is a per destination flag... > > Does that help? Hi Randy The question I have is wrt T3-timer during fast recovery. Specifically, the scenario is the following: DATA1 ----> DATA2 (lost) DATA3 ----> DATA4 ----> <---- SACK (CTSN=1) DATA5 ----> <---- SACK (CTSN=1, GAP3) <---- SACK (CTSN=1, GAP4) <---- SACK (CTSN=1, GAP5) DATA2 (fast-rtx, lost) So, the primary question is, do we reset the T3-timer when we fast retransmit DATA2. The timer would already be running from the time the first SACK arrived. Right now, lksctp does not. We changes to not retransmit data that was sent less then rto ago, we end up in a situation when the DATA2 was fast retransmitted shortly before t3-timer expired and since it was sent less then rto ago, we do not send it. As that's the only packet that would have been sent, we send nothing and as a result do not reset T3-timer and never retransmit DATA2 again. So, it makes sense to restart T3-rtx timer when fast retransmitting the lowest outstanding TSN, but there are 2 different interpretations of the point I quoted above. So, I am looking for clarification. -vlad From rrs@cisco.com Wed Apr 23 08:32:09 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id C7D073A6A74 for ; Wed, 23 Apr 2008 08:32:09 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id QAIRdja2Pi9O for ; Wed, 23 Apr 2008 08:32:08 -0700 (PDT) Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 7170A3A6B40 for ; Wed, 23 Apr 2008 08:31:28 -0700 (PDT) X-IronPort-AV: E=Sophos;i="4.25,699,1199682000"; d="scan'208";a="6110485" Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-2.cisco.com with ESMTP; 23 Apr 2008 11:31:33 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3NFVX1L012450; Wed, 23 Apr 2008 11:31:33 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3NFVWkX028577; Wed, 23 Apr 2008 15:31:32 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3NFUgKD015464 for ; Wed, 23 Apr 2008 11:30:42 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3NFUgKh015460 for sctp-impl-filtered; Wed, 23 Apr 2008 11:30:42 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f Message-Id: <480F561F.8040603@cisco.com> Date: Wed, 23 Apr 2008 08:30:39 -0700 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: Vlad Yasevich CC: SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> In-Reply-To: <480F53F6.7050406@hp.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 23 Apr 2008 15:30:39.0558 (UTC) FILETIME=[FC1CDE60:01C8A556] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=3118; t=1208964693; x=1209828693; c=relaxed/simple; s=rtpdkim1001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20 |To:=20Vlad=20Yasevich=20; bh=dnQIIyWVQzTNLQMnAlVLlqT3Bny1G2XjIVwi40n/GG8=; b=Pk+8rV4mzQGCb18LEkXlkVd81fpLqCzM3b7BomqrEtpLYYAg9Ajgpf2nln hm5jLApjMnc2BIARAhggZO0WmeO27xS30M2/fyylrwH0k0EwaIvpiFMNEZvp 5RzYaRi9t9; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=3118; t=1208964640; x=1209828640; c=relaxed/simple; s=sjdkim1004; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20; bh=dnQIIyWVQzTNLQMnAlVLlqT3Bny1G2XjIVwi40n/GG8=; b=p+n3XsmpdxzKCQBpg7VVsM0fzA9Et0/fb5CzWuahWLCdib4DRf6vyPDglb Y4KALMnvW8lLhUwZ/YD3V1gXN56zGEfUrIgND3qIPiPRMnYTPaEvAshju78g 6TIqAMGDD/VLI5HwPcdO//DQ/28ujz9xkU6ZgdBhtH678aUvXKnDY=; Authentication-Results: rtp-dkim-1; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim1001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim1004 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Vlad Yasevich wrote: > Randall Stewart wrote: >> Vlad Yasevich wrote: >>> I am trying to figure out if our implementation of fast retransmit is >>> correct and have a question about the meaning of the following: >>> >>> 4) Restart the T3-rtx timer only if the last SACK acknowledged the >>> lowest outstanding TSN number sent to that address, or the >>> endpoint is retransmitting the first outstanding DATA chunk sent >>> to that address. >>> >>> Specifically, what's is meant by the "first outstanding DATA chunk". >>> >>> Is it the very first DATA chunk that was ever sent to the >>> destination, or >>> is it the first chunk that is still outstanding? >>> >>> Thanks >>> -vlad >>> >> Basically when you recognize one or more fast retransmits need to be >> done. That first one you enter "fast recovery". >> >> During this time you do your retransmits and do NOT cut the cwnd >> if you see say another 3 strikes and mark something else. >> >> You exit fast recovery when the SACK shows up for the first >> one you recognized. >> >> The idea is you don't take more than 1 cwnd hit per window >> of data (cwnd of data that is) that has been sent. >> >> In TCP they call this fast recovery too if memory serves me correct. >> >> In our implementation in BSD when we first hit a FR we mark a flag >> that we are in FR-recovery mode. And also which TSN will exit us. >> >> When we get the ack for that, we clear the FR-Recovery flag. We also >> do not cut the cwnd on that destination. Note that FR-Recovery >> flag is a per destination flag... >> >> Does that help? > > Hi Randy > > The question I have is wrt T3-timer during fast recovery. Specifically, > the scenario is the following: > > DATA1 ----> > DATA2 (lost) > DATA3 ----> > DATA4 ----> > <---- SACK (CTSN=1) > DATA5 ----> > <---- SACK (CTSN=1, GAP3) > <---- SACK (CTSN=1, GAP4) > <---- SACK (CTSN=1, GAP5) > DATA2 (fast-rtx, lost) > > So, the primary question is, do we reset the T3-timer when we fast > retransmit > DATA2. The timer would already be running from the time the first SACK > arrived. > > Right now, lksctp does not. We changes to not retransmit data that was > sent less then > rto ago, we end up in a situation when the DATA2 was fast retransmitted > shortly before > t3-timer expired and since it was sent less then rto ago, we do not send > it. As that's > the only packet that would have been sent, we send nothing and as a > result do not > reset T3-timer and never retransmit DATA2 again. > > So, it makes sense to restart T3-rtx timer when fast retransmitting the > lowest outstanding > TSN, but there are 2 different interpretations of the point I quoted > above. So, I am > looking for clarification. Yes, restarting the T3 is appropriate in my opinion.. and it is what we do in BSD... give the association a bit more time to get all the data and get the sack's back :-) R > > -vlad > > -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From bounce@allabout.ae Sun Apr 27 02:17:13 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 42BE53A6D1E for ; Sun, 27 Apr 2008 02:17:13 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: 3.62 X-Spam-Level: *** X-Spam-Status: No, score=3.62 tagged_above=-999 required=5 tests=[BAYES_50=0.001, HTML_IMAGE_ONLY_32=1.778, HTML_IMAGE_RATIO_02=0.383, HTML_MESSAGE=0.001, MIME_HTML_ONLY=1.457] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4qtcCQK5taVj for ; Sun, 27 Apr 2008 02:17:12 -0700 (PDT) Received: from atom33.com (atom33.com [72.47.204.74]) by core3.amsl.com (Postfix) with ESMTP id 752053A6C4D for ; Sun, 27 Apr 2008 02:17:12 -0700 (PDT) Received: (qmail 7457 invoked from network); 27 Apr 2008 11:52:46 +0400 Received: from atom33.com (HELO ?72.47.204.74?) (72.47.204.74) by atom33.com with SMTP; 27 Apr 2008 11:52:46 +0400 To: From: Scorpion Dubai Reply-To: Scorpion Dubai Subject: Models, Entertainment & Films/Stills Production. Date: Sun, 27 Apr 2008 11:52:46 +0400 X-LibVersion: 3.3.2_4 MIME-Version: 1.0 Content-Type: text/html; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit X-Priority: 3 X-MSMail-Priority: Normal X-MimeOLE: Produced by SwiftMailer 3.3.2_4 X-Mailer: campaign-vault.com X-Mailer-MsgId: c2N0cC1pbXBsLWFyY2hpdmVAaWV0Zi5vcmc= X-Mailer-CSID: NDVfNjE= Message-ID: <20080427075246.12083.1211131365.swift@www.campaign-vault.com> Scorpion Dubai
 
 

--- From vladislav.yasevich@hp.com Mon Apr 28 10:33:18 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 3DACA28C2E6 for ; Mon, 28 Apr 2008 10:33:18 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -106.599 X-Spam-Level: X-Spam-Status: No, score=-106.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4, USER_IN_WHITELIST=-100] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id a6EzRXPtAmff for ; Mon, 28 Apr 2008 10:33:13 -0700 (PDT) Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 7AF1428C2EB for ; Mon, 28 Apr 2008 10:31:22 -0700 (PDT) Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-2.cisco.com with ESMTP; 28 Apr 2008 13:31:24 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3SHVOS8018581; Mon, 28 Apr 2008 13:31:24 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3SHVNVo014223; Mon, 28 Apr 2008 17:31:23 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3SHOJ5O028665 for ; Mon, 28 Apr 2008 13:24:19 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3SHOJNb028661 for sctp-impl-filtered; Mon, 28 Apr 2008 13:24:19 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to vladislav.yasevich@hp.com using -f X-From-Outside-Cisco: 15.216.28.35 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: Ah4BAFqkFUgP2BwjfGdsb2JhbACBU5AMAQELBQIGBxGYZw X-Ironport-Av: E=Sophos;i="4.25,717,1199692800"; d="scan'208";a="41925969" Message-Id: <481607CB.3050108@hp.com> Date: Mon, 28 Apr 2008 13:22:19 -0400 From: Vlad Yasevich User-Agent: Thunderbird 2.0.0.4 (X11/20070604) MIME-Version: 1.0 To: Randall Stewart CC: SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> In-Reply-To: <480F561F.8040603@cisco.com> X-Enigmail-Version: 0.95.6 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Authentication-Results: rtp-dkim-2; header.From=vladislav.yasevich@hp.com; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Hi Randy One more question about Fast-rtx and BSD implementation. It looks like you allow a chunk to be fast retransmitted multiple times if you keep getting strikes for it after the first fast retransmission. Michael pointed out to linux folks text to from 2960/4960 that says that we should fast-retransmit a given chunk only once, and we changed the code to conform the the spec. Can you point me at the paper that recommends multiple fast retransmits? Thanks -vlad From rrs@cisco.com Mon Apr 28 11:35:14 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id A28F328C1BF for ; Mon, 28 Apr 2008 11:35:14 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lIZunZ3ZkKRH for ; Mon, 28 Apr 2008 11:35:09 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id 22F0628C251 for ; Mon, 28 Apr 2008 11:35:07 -0700 (PDT) X-IronPort-AV: E=Sophos;i="4.25,717,1199682000"; d="scan'208";a="6581283" Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-1.cisco.com with ESMTP; 28 Apr 2008 14:35:09 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3SIZ9cq013789; Mon, 28 Apr 2008 14:35:09 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3SIZ834006173; Mon, 28 Apr 2008 18:35:08 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3SIWRvZ030921 for ; Mon, 28 Apr 2008 14:32:27 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3SIWRpI030917 for sctp-impl-filtered; Mon, 28 Apr 2008 14:32:27 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f Message-Id: <48161838.3040605@cisco.com> Date: Mon, 28 Apr 2008 14:32:24 -0400 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: Vlad Yasevich CC: SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> In-Reply-To: <481607CB.3050108@hp.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 28 Apr 2008 18:32:25.0240 (UTC) FILETIME=[3478A180:01C8A95E] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=1013; t=1209407709; x=1210271709; c=relaxed/simple; s=rtpdkim2001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20 |To:=20Vlad=20Yasevich=20; bh=t24hHk3AUWndLzm4k1Z/tA341rBQINKutAdKi3l2mYo=; b=UTmIjEEIw/WboX15Oa0+medkSG7wfOrjCMB6DtJUEDY4033o5C6Q0smO4H gfQ8799pKESDN9rWVpEpDXCfW9EVOetfwuwVn7+dC6+Ukpv8b+IGoUbzKwyA mcNrSaNJEw; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=1013; t=1209407545; x=1210271545; c=relaxed/simple; s=sjdkim3002; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20; bh=t24hHk3AUWndLzm4k1Z/tA341rBQINKutAdKi3l2mYo=; b=PoW4jBkQ0xtWZNy720zinopgrYASmXc1rqZCAYEdS5kRZ5JdjL+4XnGN2a gE/Q3/faii3pe1/F7aKsLb0s4za0RgUtiTuI3ovmLoumT9ziXPt4XRkQVfeL xYmDHcPfhp; Authentication-Results: rtp-dkim-2; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim2001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim3002 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Vlad Yasevich wrote: > Hi Randy > > One more question about Fast-rtx and BSD implementation. > > It looks like you allow a chunk to be fast retransmitted multiple > times if you keep getting strikes for it after the first fast > retransmission. Yes, we do that ONLY when we do NOT FR to an alternate destination. This is a switch in the code that turns that on/off > > Michael pointed out to linux folks text to from 2960/4960 that > says that we should fast-retransmit a given chunk only once, and > we changed the code to conform the the spec. > Again, you can only do multiple FR's if and only if you do NOT change the destination address when you do a FR. Otherwise it will cause problems... thats what the restriction is against.. doing a FR to an alternate address... R > Can you point me at the paper that recommends multiple fast > retransmits? > > Thanks > -vlad > -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From Michael.Tuexen@micmac.franken.de Mon Apr 28 11:53:17 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id E5E913A67A7 for ; Mon, 28 Apr 2008 11:53:17 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yIrXiVDkWTfh for ; Mon, 28 Apr 2008 11:53:17 -0700 (PDT) Received: from sj-iport-2.cisco.com (sj-iport-2.cisco.com [171.71.176.71]) by core3.amsl.com (Postfix) with ESMTP id F39313A692E for ; Mon, 28 Apr 2008 11:53:16 -0700 (PDT) Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by sj-iport-2.cisco.com with ESMTP; 28 Apr 2008 11:53:12 -0700 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3SIrCen007932; Mon, 28 Apr 2008 14:53:12 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3SIrBR6007339; Mon, 28 Apr 2008 18:53:11 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3SIpoes031471 for ; Mon, 28 Apr 2008 14:51:50 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3SIpoiI031467 for sctp-impl-filtered; Mon, 28 Apr 2008 14:51:50 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to Michael.Tuexen@micmac.franken.de using -f X-From-Outside-Cisco: 193.175.24.27 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AuABAHK5FUjBrxgbiGdsb2JhbACRXwEBAQ8gmhA X-Ironport-Av: E=Sophos;i="4.25,717,1199692800"; d="scan'208";a="41948867" CC: Randall Stewart , SCTP Implementors Message-Id: <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> From: Michael Tuexen To: Vlad Yasevich In-Reply-To: <481607CB.3050108@hp.com> Content-Type: text/plain; charset=US-ASCII; format=flowed Content-Transfer-Encoding: 7bit MIME-Version: 1.0 (Apple Message framework v919.2) Subject: Re: T3-rtx timer and fast retransmit Date: Mon, 28 Apr 2008 20:50:08 +0200 References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> X-Mailer: Apple Mail (2.919.2) Authentication-Results: rtp-dkim-1; header.From=Michael.Tuexen@micmac.franken.de; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Hi Vlad, using fast retransmitting TSN multiple times is considered in http://www.armandocaro.net/publications/2004.globecom.acaro.pdf It has not made it into RFC 4960, however it is also not forbidden to do it (There are no MUST NOT in the RFC). From my personal point of view it does not make sense to multiply fast retransmit a TSN within a time which corresponds to the RTO calculation before doing the [RTO.Min, RTO.Max] computation. Retransmitting earlier looks like forward error correction, since you can not expect a SACK to be received before RTT. Best regards Michael On Apr 28, 2008, at 7:22 PM, Vlad Yasevich wrote: > Hi Randy > > One more question about Fast-rtx and BSD implementation. > > It looks like you allow a chunk to be fast retransmitted multiple > times if you keep getting strikes for it after the first fast > retransmission. > > Michael pointed out to linux folks text to from 2960/4960 that > says that we should fast-retransmit a given chunk only once, and > we changed the code to conform the the spec. > > Can you point me at the paper that recommends multiple fast > retransmits? > > Thanks > -vlad > From rrs@cisco.com Mon Apr 28 13:33:52 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 5EF4D3A6B0C for ; Mon, 28 Apr 2008 13:33:52 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 5ugdcBsjWkfx for ; Mon, 28 Apr 2008 13:33:51 -0700 (PDT) Received: from rtp-iport-2.cisco.com (rtp-iport-2.cisco.com [64.102.122.149]) by core3.amsl.com (Postfix) with ESMTP id 137913A6860 for ; Mon, 28 Apr 2008 13:33:50 -0700 (PDT) X-IronPort-AV: E=Sophos;i="4.25,718,1199682000"; d="scan'208";a="6600233" Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-2.cisco.com with ESMTP; 28 Apr 2008 16:33:53 -0400 Received: from rtp-core-2.cisco.com (rtp-core-2.cisco.com [64.102.124.13]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3SKXrMg007169; Mon, 28 Apr 2008 16:33:53 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-2.cisco.com (8.13.8/8.13.8) with ESMTP id m3SKXqcX019760; Mon, 28 Apr 2008 20:33:53 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3SKVNJG002482 for ; Mon, 28 Apr 2008 16:31:23 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3SKVNfR002478 for sctp-impl-filtered; Mon, 28 Apr 2008 16:31:23 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f Message-Id: <48163417.9090908@cisco.com> Date: Mon, 28 Apr 2008 16:31:19 -0400 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: Michael Tuexen CC: Vlad Yasevich , SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> In-Reply-To: <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 28 Apr 2008 20:31:20.0437 (UTC) FILETIME=[D1614A50:01C8A96E] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=2025; t=1209414833; x=1210278833; c=relaxed/simple; s=rtpdkim2001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20 |To:=20Michael=20Tuexen=20; bh=aYj5fKpNNhP59AM1TwRX0zqi4vfOcnDCj/PQeSouKas=; b=UKtCVS+WS5Wknoj/wYHBHim8IsL5R9CHFzjKc7QhhFs16EE8T3qPW4COP/ wKdqv0JFAPxQPo5nZMxW4+zztnUHBXzbz96LJJvxQlgWxP1LqoQC6UzMsGoj NleFYdEotG; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=2025; t=1209414681; x=1210278681; c=relaxed/simple; s=sjdkim4002; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20; bh=aYj5fKpNNhP59AM1TwRX0zqi4vfOcnDCj/PQeSouKas=; b=sKijtNbxspJhTyuhh4uSsJP4BwnzZ/ygY4+PWC51eej9UG56aqXJCz3RqO pzOArzD++1Q4uyVmUukRdzjB35sstzHGZVSfS3uid41jtBE72jKJyNye+JHC x6f+SpX/XS; Authentication-Results: rtp-dkim-2; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim2001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim4002 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Michael: There are ways to do this without worrying about the time.. What we do is: A) Store the TSN where strikes count.. so lets say we send TSN 1... we store 1 with it. B) When we hit the FR when SACK saying cum=0, gap 2-4, we do the FR but we also record the next to send TSN with TSN1, so if the next one was 6, we mark that in TSN1. C) Now when we get a gap ack the TSN must be greater than 6 before we can consider striking it.. Again when we reach 3 strikes.. ACK of TSN 9, we do a subsequent FR. This only is enabled of course if we are not sending a FR to the alternate destination (if multi-homed).. R Michael Tuexen wrote: > Hi Vlad, > > using fast retransmitting TSN multiple times is considered in > http://www.armandocaro.net/publications/2004.globecom.acaro.pdf > It has not made it into RFC 4960, however it is also not forbidden > to do it (There are no MUST NOT in the RFC). > > From my personal point of view it does not make sense to multiply > fast retransmit a TSN within a time which corresponds to the RTO > calculation before doing the [RTO.Min, RTO.Max] computation. > Retransmitting earlier looks like forward error correction, since > you can not expect a SACK to be received before RTT. > > Best regards > Michael > > On Apr 28, 2008, at 7:22 PM, Vlad Yasevich wrote: > >> Hi Randy >> >> One more question about Fast-rtx and BSD implementation. >> >> It looks like you allow a chunk to be fast retransmitted multiple >> times if you keep getting strikes for it after the first fast >> retransmission. >> >> Michael pointed out to linux folks text to from 2960/4960 that >> says that we should fast-retransmit a given chunk only once, and >> we changed the code to conform the the spec. >> >> Can you point me at the paper that recommends multiple fast >> retransmits? >> >> Thanks >> -vlad >> > > -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From Michael.Tuexen@micmac.franken.de Tue Apr 29 08:15:58 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 7839428C34F for ; Tue, 29 Apr 2008 08:15:58 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.599 X-Spam-Level: X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id UMnvQKvSXtNi for ; Tue, 29 Apr 2008 08:15:52 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id D859928C32B for ; Tue, 29 Apr 2008 08:15:51 -0700 (PDT) Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-1.cisco.com with ESMTP; 29 Apr 2008 11:15:52 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3TFFqtw006353; Tue, 29 Apr 2008 11:15:52 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3TFFp1j024957; Tue, 29 Apr 2008 15:15:52 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3TFCnAI014780 for ; Tue, 29 Apr 2008 11:12:49 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3TFCnTt014776 for sctp-impl-filtered; Tue, 29 Apr 2008 11:12:49 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to Michael.Tuexen@micmac.franken.de using -f X-From-Outside-Cisco: 193.175.24.27 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AjkCAGjXFkjBrxgbiGdsb2JhbACRTQEBAQ8gmmk X-Ironport-Av: E=Sophos;i="4.25,723,1199692800"; d="scan'208";a="42254503" CC: Vlad Yasevich , SCTP Implementors Message-Id: <2BB52381-F441-495D-8983-BC98469AE8CB@micmac.franken.de> From: Michael Tuexen To: Randall Stewart In-Reply-To: <48163417.9090908@cisco.com> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes Content-Transfer-Encoding: 7bit MIME-Version: 1.0 (Apple Message framework v919.2) Subject: Re: T3-rtx timer and fast retransmit Date: Tue, 29 Apr 2008 17:10:55 +0200 References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> <48163417.9090908@cisco.com> X-Mailer: Apple Mail (2.919.2) Authentication-Results: rtp-dkim-1; header.From=Michael.Tuexen@micmac.franken.de; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Hi Randy, yes, that does make sure that you do not fast retransmit before it can be received. You can base that on TSN, not on time as I indicated. This procedure is not blindly doing multiple fast retransmits on every three gap reports and is not specified in any RFC. It is a BSD specific extension. Best regards Michael On Apr 28, 2008, at 10:31 PM, Randall Stewart wrote: > Michael: > > There are ways to do this without worrying about the time.. > > What we do is: > > A) Store the TSN where strikes count.. so lets say we > send TSN 1... we store 1 with it. > > B) When we hit the FR when SACK saying cum=0, gap 2-4, we > do the FR but we also record the next to send TSN with TSN1, > so if the next one was 6, we mark that in TSN1. > > C) Now when we get a gap ack the TSN must be greater than 6 before > we can consider striking it.. Again when we reach 3 strikes.. > ACK of > TSN 9, we do a subsequent FR. > > This only is enabled of course if we are not sending a FR to the > alternate > destination (if multi-homed).. > > R > > > Michael Tuexen wrote: >> Hi Vlad, >> using fast retransmitting TSN multiple times is considered in >> http://www.armandocaro.net/publications/2004.globecom.acaro.pdf >> It has not made it into RFC 4960, however it is also not forbidden >> to do it (There are no MUST NOT in the RFC). >> From my personal point of view it does not make sense to multiply >> fast retransmit a TSN within a time which corresponds to the RTO >> calculation before doing the [RTO.Min, RTO.Max] computation. >> Retransmitting earlier looks like forward error correction, since >> you can not expect a SACK to be received before RTT. >> Best regards >> Michael >> On Apr 28, 2008, at 7:22 PM, Vlad Yasevich wrote: >>> Hi Randy >>> >>> One more question about Fast-rtx and BSD implementation. >>> >>> It looks like you allow a chunk to be fast retransmitted multiple >>> times if you keep getting strikes for it after the first fast >>> retransmission. >>> >>> Michael pointed out to linux folks text to from 2960/4960 that >>> says that we should fast-retransmit a given chunk only once, and >>> we changed the code to conform the the spec. >>> >>> Can you point me at the paper that recommends multiple fast >>> retransmits? >>> >>> Thanks >>> -vlad >>> > > > -- > Randall Stewart > NSSTG - Cisco Systems Inc. > 803-345-0369 803-317-4952 (cell) > From rrs@cisco.com Tue Apr 29 13:57:12 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 848CF28C0FC for ; Tue, 29 Apr 2008 13:57:12 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.299 X-Spam-Level: X-Spam-Status: No, score=-6.299 tagged_above=-999 required=5 tests=[AWL=-0.300, BAYES_00=-2.599, J_CHICKENPOX_44=0.6, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6BiQgjNwkv0L for ; Tue, 29 Apr 2008 13:57:11 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id 321743A6B99 for ; Tue, 29 Apr 2008 13:57:11 -0700 (PDT) Received: from rtp-dkim-2.cisco.com ([64.102.121.159]) by rtp-iport-1.cisco.com with ESMTP; 29 Apr 2008 16:57:13 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-2.cisco.com (8.12.11/8.12.11) with ESMTP id m3TKvDi6010957; Tue, 29 Apr 2008 16:57:13 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3TKvCec017226; Tue, 29 Apr 2008 20:57:13 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3TKqPV6026215 for ; Tue, 29 Apr 2008 16:52:25 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3TKqPZm026211 for sctp-impl-filtered; Tue, 29 Apr 2008 16:52:25 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to rrs@cisco.com using -f X-Ironport-Av: E=Sophos;i="4.25,724,1199692800"; d="scan'208";a="24483803" Message-Id: <48178A85.5040409@cisco.com> Date: Tue, 29 Apr 2008 16:52:21 -0400 From: Randall Stewart User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8 MIME-Version: 1.0 To: Michael Tuexen CC: Vlad Yasevich , SCTP Implementors Subject: Re: T3-rtx timer and fast retransmit References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> <48163417.9090908@cisco.com> <2BB52381-F441-495D-8983-BC98469AE8CB@micmac.franken.de> In-Reply-To: <2BB52381-F441-495D-8983-BC98469AE8CB@micmac.franken.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Originalarrivaltime: 29 Apr 2008 20:52:22.0759 (UTC) FILETIME=[EC321770:01C8AA3A] DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=2831; t=1209502633; x=1210366633; c=relaxed/simple; s=rtpdkim2001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20 |To:=20Michael=20Tuexen=20; bh=nfGo0IsPwred9zF6HYRrGXvm+dSMvM6XCUottArp2eE=; b=dwRRO/BEZp7cgapn5HiF1UOJIFGw7fHnqBLuDPouTY2nAwd+3SRAviHH0g 85NYOXKybf/vzmtDyXo2MbwpSsj10syThtIaM+0X/oykK63Z/gBB2OAJuJRp 1rfXTsIrDe; Dkim-Signature: v=1; a=rsa-sha256; q=dns/txt; l=2831; t=1209502343; x=1210366343; c=relaxed/simple; s=sjdkim4002; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=rrs@cisco.com; z=From:=20Randall=20Stewart=20 |Subject:=20Re=3A=20T3-rtx=20timer=20and=20fast=20retransmi t |Sender:=20; bh=nfGo0IsPwred9zF6HYRrGXvm+dSMvM6XCUottArp2eE=; b=mwqi9A+RVBXdC62X+K06E+GNFesbZ+0tGjYU7UfctrnwFbB9/3oNbZiE7o CdOwJa9BCmRKr1kpq2R1qqdc3YVIyedjDrmeIn9zOkjLgeVBUzoCNrdXol8H UyL/REb4JE; Authentication-Results: rtp-dkim-2; header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/rtpdkim2001 verified; ); header.From=rrs@cisco.com; dkim=pass ( sig from cisco.com/sjdkim4002 verified; ); X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 Michael Tuexen wrote: > Hi Randy, > > yes, that does make sure that you do not fast retransmit > before it can be received. You can base that on TSN, not > on time as I indicated. > > This procedure is not blindly doing multiple fast retransmits > on every three gap reports and is not specified in any RFC. It > is a BSD specific extension. And also documented in several papers that Armando, Jana,Paul and I worked on I think.. cant remember precisely which one though :-) R > > Best regards > Michael > > On Apr 28, 2008, at 10:31 PM, Randall Stewart wrote: > >> Michael: >> >> There are ways to do this without worrying about the time.. >> >> What we do is: >> >> A) Store the TSN where strikes count.. so lets say we >> send TSN 1... we store 1 with it. >> >> B) When we hit the FR when SACK saying cum=0, gap 2-4, we >> do the FR but we also record the next to send TSN with TSN1, >> so if the next one was 6, we mark that in TSN1. >> >> C) Now when we get a gap ack the TSN must be greater than 6 before >> we can consider striking it.. Again when we reach 3 strikes.. ACK of >> TSN 9, we do a subsequent FR. >> >> This only is enabled of course if we are not sending a FR to the >> alternate >> destination (if multi-homed).. >> >> R >> >> >> Michael Tuexen wrote: >>> Hi Vlad, >>> using fast retransmitting TSN multiple times is considered in >>> http://www.armandocaro.net/publications/2004.globecom.acaro.pdf >>> It has not made it into RFC 4960, however it is also not forbidden >>> to do it (There are no MUST NOT in the RFC). >>> From my personal point of view it does not make sense to multiply >>> fast retransmit a TSN within a time which corresponds to the RTO >>> calculation before doing the [RTO.Min, RTO.Max] computation. >>> Retransmitting earlier looks like forward error correction, since >>> you can not expect a SACK to be received before RTT. >>> Best regards >>> Michael >>> On Apr 28, 2008, at 7:22 PM, Vlad Yasevich wrote: >>>> Hi Randy >>>> >>>> One more question about Fast-rtx and BSD implementation. >>>> >>>> It looks like you allow a chunk to be fast retransmitted multiple >>>> times if you keep getting strikes for it after the first fast >>>> retransmission. >>>> >>>> Michael pointed out to linux folks text to from 2960/4960 that >>>> says that we should fast-retransmit a given chunk only once, and >>>> we changed the code to conform the the spec. >>>> >>>> Can you point me at the paper that recommends multiple fast >>>> retransmits? >>>> >>>> Thanks >>>> -vlad >>>> >> >> >> -- >> Randall Stewart >> NSSTG - Cisco Systems Inc. >> 803-345-0369 803-317-4952 (cell) >> > > -- Randall Stewart NSSTG - Cisco Systems Inc. 803-345-0369 803-317-4952 (cell) From Michael.Tuexen@micmac.franken.de Tue Apr 29 23:26:34 2008 Return-Path: X-Original-To: ietfarch-sctp-impl-archive@core3.amsl.com Delivered-To: ietfarch-sctp-impl-archive@core3.amsl.com Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 5905B3A677E for ; Tue, 29 Apr 2008 23:26:34 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -6.299 X-Spam-Level: X-Spam-Status: No, score=-6.299 tagged_above=-999 required=5 tests=[AWL=-0.300, BAYES_00=-2.599, J_CHICKENPOX_44=0.6, RCVD_IN_DNSWL_MED=-4] Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id COg792glufyb for ; Tue, 29 Apr 2008 23:26:33 -0700 (PDT) Received: from rtp-iport-1.cisco.com (rtp-iport-1.cisco.com [64.102.122.148]) by core3.amsl.com (Postfix) with ESMTP id D58973A6E05 for ; Tue, 29 Apr 2008 23:26:02 -0700 (PDT) Received: from rtp-dkim-1.cisco.com ([64.102.121.158]) by rtp-iport-1.cisco.com with ESMTP; 30 Apr 2008 02:26:05 -0400 Received: from rtp-core-1.cisco.com (rtp-core-1.cisco.com [64.102.124.12]) by rtp-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m3U6Q3nB009146; Wed, 30 Apr 2008 02:26:03 -0400 Received: from mailguard.cisco.com (mailguard.cisco.com [64.102.28.79]) by rtp-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m3U6Q2Dg016159; Wed, 30 Apr 2008 06:26:02 GMT Received: from mailguard.cisco.com (localhost.cisco.com [127.0.0.1]) by mailguard.cisco.com (8.12.11/8.13.4) with ESMTP id m3U6P8ip013985 for ; Wed, 30 Apr 2008 02:25:08 -0400 Received: (from mailnull@localhost) by mailguard.cisco.com (8.12.11/8.12.11/Submit) id m3U6P8ea013981 for sctp-impl-filtered; Wed, 30 Apr 2008 02:25:08 -0400 X-Authentication-Warning: mailguard.cisco.com: mailnull set sender to Michael.Tuexen@micmac.franken.de using -f X-From-Outside-Cisco: 193.175.24.27 X-Ironport-Anti-Spam-Filtered: true X-Ironport-Anti-Spam-Result: AqUBAK+tF0jBrxgbiGdsb2JhbACRUQEBAQ8gmw8 X-Ironport-Av: E=Sophos;i="4.25,726,1199692800"; d="scan'208";a="29446789" CC: Vlad Yasevich , SCTP Implementors Message-Id: From: Michael Tuexen To: Randall Stewart In-Reply-To: <48178A85.5040409@cisco.com> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes Content-Transfer-Encoding: 7bit MIME-Version: 1.0 (Apple Message framework v919.2) Subject: Re: T3-rtx timer and fast retransmit Date: Wed, 30 Apr 2008 08:23:22 +0200 References: <480E03DE.9090501@hp.com> <480F4907.9010506@cisco.com> <480F53F6.7050406@hp.com> <480F561F.8040603@cisco.com> <481607CB.3050108@hp.com> <3A3FDBDB-C4A3-4C8A-8283-801B11DB0E51@micmac.franken.de> <48163417.9090908@cisco.com> <2BB52381-F441-495D-8983-BC98469AE8CB@micmac.franken.de> <48178A85.5040409@cisco.com> X-Mailer: Apple Mail (2.919.2) Authentication-Results: rtp-dkim-1; header.From=Michael.Tuexen@micmac.franken.de; dkim=neutral X-Mailguard-Protected-List: sctp-impl X-Mailguard-Version: 1.2 On Apr 29, 2008, at 10:52 PM, Randall Stewart wrote: > Michael Tuexen wrote: >> Hi Randy, >> yes, that does make sure that you do not fast retransmit >> before it can be received. You can base that on TSN, not >> on time as I indicated. >> This procedure is not blindly doing multiple fast retransmits >> on every three gap reports and is not specified in any RFC. It >> is a BSD specific extension. > > And also documented in several papers that Armando, Jana,Paul > and I worked on I think.. cant remember precisely which one though :-) True, I send a link to the paper a couple of days ago to the list... Best regards Michael > > > R > >> Best regards >> Michael >> On Apr 28, 2008, at 10:31 PM, Randall Stewart wrote: >>> Michael: >>> >>> There are ways to do this without worrying about the time.. >>> >>> What we do is: >>> >>> A) Store the TSN where strikes count.. so lets say we >>> send TSN 1... we store 1 with it. >>> >>> B) When we hit the FR when SACK saying cum=0, gap 2-4, we >>> do the FR but we also record the next to send TSN with TSN1, >>> so if the next one was 6, we mark that in TSN1. >>> >>> C) Now when we get a gap ack the TSN must be greater than 6 before >>> we can consider striking it.. Again when we reach 3 strikes.. >>> ACK of >>> TSN 9, we do a subsequent FR. >>> >>> This only is enabled of course if we are not sending a FR to the >>> alternate >>> destination (if multi-homed).. >>> >>> R >>> >>> >>> Michael Tuexen wrote: >>>> Hi Vlad, >>>> using fast retransmitting TSN multiple times is considered in >>>> http://www.armandocaro.net/publications/2004.globecom.acaro.pdf >>>> It has not made it into RFC 4960, however it is also not forbidden >>>> to do it (There are no MUST NOT in the RFC). >>>> From my personal point of view it does not make sense to multiply >>>> fast retransmit a TSN within a time which corresponds to the RTO >>>> calculation before doing the [RTO.Min, RTO.Max] computation. >>>> Retransmitting earlier looks like forward error correction, since >>>> you can not expect a SACK to be received before RTT. >>>> Best regards >>>> Michael >>>> On Apr 28, 2008, at 7:22 PM, Vlad Yasevich wrote: >>>>> Hi Randy >>>>> >>>>> One more question about Fast-rtx and BSD implementation. >>>>> >>>>> It looks like you allow a chunk to be fast retransmitted multiple >>>>> times if you keep getting strikes for it after the first fast >>>>> retransmission. >>>>> >>>>> Michael pointed out to linux folks text to from 2960/4960 that >>>>> says that we should fast-retransmit a given chunk only once, and >>>>> we changed the code to conform the the spec. >>>>> >>>>> Can you point me at the paper that recommends multiple fast >>>>> retransmits? >>>>> >>>>> Thanks >>>>> -vlad >>>>> >>> >>> >>> -- >>> Randall Stewart >>> NSSTG - Cisco Systems Inc. >>> 803-345-0369 803-317-4952 (cell) >>> > > > -- > Randall Stewart > NSSTG - Cisco Systems Inc. > 803-345-0369 803-317-4952 (cell) >