From owner-freebsd-net Sun Oct 24 17: 7:20 1999 Delivered-To: freebsd-net@freebsd.org Received: from cc.weber.edu (cc.weber.edu [137.190.1.2]) by hub.freebsd.org (Postfix) with ESMTP id 15FA015145 for ; Sun, 24 Oct 1999 17:07:18 -0700 (PDT) (envelope-from jlewis@cc.weber.edu) Received: from offcampus.weber.edu ([137.190.3.223]) by cc.WEBER.EDU (PMDF V5.1-12 #7039) with SMTP id <01JHISR7O4V48Y67WP@cc.WEBER.EDU> for freebsd-net@FreeBSD.ORG; Sun, 24 Oct 1999 18:07:17 MST Date: Sun, 24 Oct 1999 18:07:19 -0600 (Mountain Daylight Time) From: Jason Lewis Subject: Re: ppp In-reply-to: <199910071026.LAA04197@hak.lan.Awfulhak.org> To: Brian Somers Cc: freebsd-net@FreeBSD.ORG Message-id: MIME-version: 1.0 Content-type: TEXT/PLAIN; charset=US-ASCII Content-transfer-encoding: 7BIT X-X-Sender: jlewis9@gwmta1.weber.edu Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Ok. I tried this and found out that my router is intermittant. I suspect that I need more ram than 8 Mb. I could not find anything on how much is needed with a router and ppp. I think I need atleast 32M for the box to work with routing. Any insite on this? On Thu, 7 Oct 1999, Brian Somers wrote: > > How can one route public IP's over ppp without IP Masquerading? I tried > > enabling forwarding, but it didnot work. Do I need to use pppd instead of > > ppp? > > I'd advise using tcpdump (or enabling ppps tcp/ip logging) at each > point of the packets journey to discover what's not forwarding it... > > Ppp is functionally the same as pppd in this respect. > > -- > Brian > > Don't _EVER_ lose your sense of humour ! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 1:33:26 1999 Delivered-To: freebsd-net@freebsd.org Received: from doughnut.cc.uq.edu.au (doughnut.cc.uq.edu.au [130.102.128.239]) by hub.freebsd.org (Postfix) with ESMTP id 0836C14BFC; Mon, 25 Oct 1999 01:33:13 -0700 (PDT) (envelope-from T.Jago@its.uq.edu.au) Received: (from tony@localhost) by doughnut.cc.uq.edu.au (8.9.3/8.9.3) id SAA55046; Mon, 25 Oct 1999 18:33:07 +1000 (EST) (envelope-from T.Jago@its.uq.edu.au) X-Authentication-Warning: doughnut.cc.uq.edu.au: tony set sender to T.Jago@its.uq.edu.au using -f Date: Mon, 25 Oct 1999 18:33:06 +1000 (EST) X-Sender: tony@doughnut.cc.uq.edu.au To: Julian Elischer Cc: freebsd-net@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Very Poor Samba -> Win9x performance [more] In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII From: Tony Jago Organization: University of Queensland X-Key-Fingerprint: 48 3F 5D FB 37 51 01 C5 A1 82 B1 6B 76 2C 75 9D X-PGP-Key-URL: http://doughnut.cc.uq.edu.au/pub/pgp-keys/T.Jago@its.uq.edu.au X-Direct-Email: tony@scar.uq.edu.au Comments: This mail has be en PGP signed. (http://www.pgp.com/) Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Further to my problem with slow smbclient "put"'s. 1. FreeBSD 3.[23] machine -> Win98 slow (9k per second). 2. Solaris machine -> Win98 fast (500+k per second). 3. FreeBSD 2.2.7 machine -> Win98 fast (500+k per second). Just incase I was doing something insane, I installed a brand new 3.3-RELEASE machine running the GENERIC kernel, installed samba from the ports and did a smbclient "put" - still very slow performance. This is very repeatable, I have tryed this on 3 different hardware platforms on different networks and different network cards, they all do exactly the same thing. Tweeking with the samba config file makes no difference, this must be something way beyond that. We are talking 9k a second, not much more then a modem out of 100M ethernet cards on switched networks! These machine can ftp at rates closer to 5 megabytes a second and above. smbclient \\\\win98box\\test -N -c "put 1mfile" Could some reading this that has samba installed on a FreeBSD 3.3 machine please try this just so that I know that I am not going insane? Thanks Tony --- Tony Jago, System Administrator, E-Mail: T.Jago@its.uq.edu.au Server and Security Group, Phone: +61 7 3365 4078 Information Technology Services, The University of Queensland. Brisbane, Australia. 4072. > On Sun, 24 Oct 1999, Tony Jago wrote: > > > > > Hello, when transferring a file from a FreeBSD box to a Win9x share using > > smbclient the performance appears to be very slow (ie. about 10k per > > second). I am running FreeBSD 3.3-STABLE (as of 24/10/99) and Samba > > 2.0.5a although the problem occurs on slightly older versions of FreeBSD > > and of samba. > > > > As you can see, the get and put performance is vastly different. The > > tests were performed on a dedicated network. > > > > # smbclient \\\\panic\\upload -N -c "put 1mb.dat" > > Added interface ip=10.0.2.1 bcast=10.0.2.255 nmask=255.255.255.0 > > Got a positive name query response from 10.0.2.18 ( 10.0.2.18 ) > > putting file 1mb.dat as \1mb.dat (9.82971 kb/s) (average 9.82971 kb/s) > > > > # smbclient \\\\panic\\upload -N -c "get 1mb.dat" > > Added interface ip=10.0.2.1 bcast=10.0.2.255 nmask=255.255.255.0 > > Got a positive name query response from 10.0.2.18 ( 10.0.2.18 ) > > getting file 1mb.dat of size 1048576 as 1mb.dat (425.603 kb/s) (average > > 425.603 kb/s) > > > > If I boot the FreeBSD box into Windows then it can transfer files to the > > other windows box at high speed. > > > > I have reproduced the problem on 3 different FreeBSD boxes and 2 > > different windows boxes (win95 & win98). > > > > I am unsure if the problem is a TCP problem or a Samba problem. During > > the slow transfer, netstat always reports 2111 bytes in the SendQ on the > > BSD box. > > > > # netstat -n > > Active Internet connections > > Proto Recv-Q Send-Q Local Address Foreign Address (state) > > tcp 0 2111 10.0.2.1.1108 10.0.2.18.139 ESTABLISHED > > > > A tcpdump of the transfer however shows some longish pauses waiting for > > the windows box to reply. This pause seems to be about the same length of > > time no matter if the windows box is is a Pentium 120 or a PII 333. > > > > 10.0.2.1 (FreeBSD PC) 10.0.2.18 (Windows 95 PC) > > > > # tcpdump -i ed0 -n > > tcpdump: listening on ed0 > > 14:22:47.143804 10.0.2.1.1282 > 10.0.2.255.137: udp 50 > > 14:22:47.144432 10.0.2.18.137 > 10.0.2.1.1282: udp 62 > > 14:22:47.444199 10.0.2.1.1109 > 10.0.2.18.139: S 811544824:811544824(0) > > win 16384 (DF) > > 14:22:47.444710 10.0.2.18.139 > 10.0.2.1.1109: S 17659612:17659612(0) ack > > 811544825 win 8760 (DF) > > 14:22:47.444921 10.0.2.1.1109 > 10.0.2.18.139: . ack 1 win 17520 (DF) > > 14:22:47.704063 10.0.2.1.1109 > 10.0.2.18.139: P 1:77(76) ack 1 win 17520 > > (DF) > > 14:22:47.704723 10.0.2.18.139 > 10.0.2.1.1109: P 1:5(4) ack 77 win 8684 > > (DF) > > 14:22:47.705465 10.0.2.1.1109 > 10.0.2.18.139: P 77:245(168) ack 5 win > > 17520 (DF) > > 14:22:47.706348 10.0.2.18.139 > 10.0.2.1.1109: P 5:86(81) ack 245 win 8516 > > (DF) > > 14:22:47.718935 10.0.2.1.1109 > 10.0.2.18.139: P 245:339(94) ack 86 win > > 17520 (DF) > > 14:22:47.719783 10.0.2.18.139 > 10.0.2.1.1109: P 86:131(45) ack 339 win > > 8422 (DF) > > 14:22:47.726618 10.0.2.1.1109 > 10.0.2.18.139: P 339:431(92) ack 131 win > > 17520 (DF) > > 14:22:47.728456 10.0.2.18.139 > 10.0.2.1.1109: P 131:177(46) ack 431 win > > 8330 (DF) > > 14:22:47.729473 10.0.2.1.1109 > 10.0.2.18.139: P 431:509(78) ack 177 win > > 17520 (DF) > > 14:22:47.732845 10.0.2.18.139 > 10.0.2.1.1109: P 177:246(69) ack 509 win > > 8252 (DF) > > 14:22:47.740104 10.0.2.1.1109 > 10.0.2.18.139: . 509:1969(1460) ack 246 > > win 17520 (DF) > > 14:22:47.933404 10.0.2.18.139 > 10.0.2.1.1109: . ack 1969 win 8760 (DF) > > 14:22:47.933898 10.0.2.1.1109 > 10.0.2.18.139: P 1969:2620(651) ack 246 > > win 17520 (DF) > > 14:22:47.935734 10.0.2.18.139 > 10.0.2.1.1109: P 246:297(51) ack 2620 win > > 8109 (DF) > > 14:22:47.936794 10.0.2.1.1109 > 10.0.2.18.139: . 2620:4080(1460) ack 297 > > win 17520 (DF) > > 14:22:48.135970 10.0.2.18.139 > 10.0.2.1.1109: . ack 4080 win 8760 (DF) > > 14:22:48.136422 10.0.2.1.1109 > 10.0.2.18.139: P 4080:4731(651) ack 297 > > win 17520 (DF) > > 14:22:48.138145 10.0.2.18.139 > 10.0.2.1.1109: P 297:348(51) ack 4731 win > > 8109 (DF) > > 14:22:48.139198 10.0.2.1.1109 > 10.0.2.18.139: . 4731:6191(1460) ack 348 > > win 17520 (DF) > > 14:22:48.338442 10.0.2.18.139 > 10.0.2.1.1109: . ack 6191 win 8760 (DF) > > 14:22:48.338894 10.0.2.1.1109 > 10.0.2.18.139: P 6191:6842(651) ack 348 > > win 17520 (DF) > > 14:22:48.340633 10.0.2.18.139 > 10.0.2.1.1109: P 348:399(51) ack 6842 win > > 8109 (DF) > > 14:22:48.341720 10.0.2.1.1109 > 10.0.2.18.139: . 6842:8302(1460) ack 399 > > win 17520 (DF) > > 14:22:48.540938 10.0.2.18.139 > 10.0.2.1.1109: . ack 8302 win 8760 (DF) > > 14:22:48.541398 10.0.2.1.1109 > 10.0.2.18.139: P 8302:8953(651) ack 399 > > win 17520 (DF) > > 14:22:48.543140 10.0.2.18.139 > 10.0.2.1.1109: P 399:450(51) ack 8953 win > > 8109 (DF) > > 14:22:48.544242 10.0.2.1.1109 > 10.0.2.18.139: . 8953:10413(1460) ack 450 > > win 17520 (DF) > > > > I do have "tcp_extensions" switched on but I have tried putting them off > > and tweaking with the Microsoft TCP stack as well and it makes no > > difference. Samba has "TCP_NODELAY" as a socket option. > > > > If anybody can shed some light on this problem it would be great. Thanks > > in Advance, > > > > Tony > > > > --- > > Tony Jago, System Administrator, E-Mail: T.Jago@its.uq.edu.au > > Server and Security Group, Phone: +61 7 3365 4078 > > Information Technology Services, > > The University of Queensland. Brisbane, Australia. 4072. > > > > > > > > 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 From owner-freebsd-net Mon Oct 25 4: 5:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp-out2.bellatlantic.net (smtp-out2.bellatlantic.net [199.45.39.157]) by hub.freebsd.org (Postfix) with ESMTP id 355FB1518C for ; Mon, 25 Oct 1999 04:05:43 -0700 (PDT) (envelope-from walterr@bellatlantic.net) Received: from bellatlantic.net (adsl-151-202-19-237.bellatlantic.net [151.202.19.237]) by smtp-out2.bellatlantic.net (8.9.1/8.9.1) with ESMTP id HAA16658; Mon, 25 Oct 1999 07:10:53 -0400 (EDT) Message-ID: <38143985.4ADE89A5@bellatlantic.net> Date: Mon, 25 Oct 1999 07:05:41 -0400 From: Tony Rentschler X-Mailer: Mozilla 4.7 [en] (X11; U; Linux 2.0.36 i386) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-net@freebsd.org Subject: VPN with PPTP or PPP Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'm trying to set up a pptpclient (from the ports) on my FreeBSD 3.3 machine at home so I can connect to my company's VPN through my ISP, which is BellAtlantic and which I access through an ADSL line. I've got a static IP address and my machine is set up accordingly to connect to BA on xl0, my 3COM NIC. The VPN at work server is running Windows NT. I can get this to work from Windows98 (my machine is dual-boot). I've looked at the ppp.conf sample files, and the pptpclient sample files, and, believe me, I've scoured deja.com for clues, but I just can't seem to get my setup to work. Who's coming in to a Windows-hosted VPN from a FreeBSD box over an ADSL or cable-modem? Can you help? Thanks, Tony To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 6: 5:49 1999 Delivered-To: freebsd-net@freebsd.org Received: from timingpdc.timing.com (timingpdc.timing.com [206.168.13.194]) by hub.freebsd.org (Postfix) with ESMTP id 1743214CC2; Mon, 25 Oct 1999 06:05:45 -0700 (PDT) (envelope-from jhein@timing.com) Received: from Daffy.timing.com ([206.168.13.218]) by timingpdc.timing.com (Post.Office MTA v3.1.2 release (PO205-101c) ID# 103-49575U100L2S100) with ESMTP id AAA331; Mon, 25 Oct 1999 07:06:23 -0600 Received: (from jhein@localhost) by Daffy.timing.com (8.9.3/8.9.3) id HAA28016; Mon, 25 Oct 1999 07:05:36 -0600 (MDT) (envelope-from jhein@Daffy.timing.com) X-Authentication-Warning: Daffy.timing.com: jhein set sender to jhein@Daffy.timing.com using -f MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14356.21920.670949.649870@Daffy.timing.com> Date: Mon, 25 Oct 1999 07:05:36 -0600 (MDT) From: "John E. Hein" To: Tony Jago Cc: Julian Elischer , freebsd-net@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Very Poor Samba -> Win9x performance [more] In-Reply-To: References: X-Mailer: VM 6.75 under Emacs 20.4.1 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Tony Jago wrote at 18:33 +1000 on Oct 25: > Further to my problem with slow smbclient "put"'s. > > 1. FreeBSD 3.[23] machine -> Win98 slow (9k per second). > 2. Solaris machine -> Win98 fast (500+k per second). > 3. FreeBSD 2.2.7 machine -> Win98 fast (500+k per second). > > Just incase I was doing something insane, I installed a brand new > 3.3-RELEASE machine running the GENERIC kernel, installed samba from the > ports and did a smbclient "put" - still very slow performance. Make sure you've CVSup'd the latest samba port. On 9/24 patch-ah was added which makes a mod to lib/util_sock.c to resolve this problem on FreeBSD machines: Here's the log description of the patch. It seems to explain exactly your problem, but you didn't give enough info about your samba port installation to tell whether you got this patch or not. ----- start log msg for patch-ah ----- revision 1.1 date: 1999/09/24 04:29:42; author: cpiazza; state: Exp; Add a patch to fix a problem with very low write speeds with samba on freebsd. Following is description made by one of the Samba developers: "The problem is we switched to using recv() with the MSG_WAITALL flag instead of read(). This makes Samba faster on most systems. On FreeBSD it causes a massive slowdown and I don't know why. To fix it, change the definition of MSG_WAITALL to be zero in lib/util_sock.c and recompile 2.0.5a. I'm hoping someone from FreeBSD will get back to me with some explanation. Regards, Jeremy Allison, Samba Team." This problem where observed by almost all samba-2.0.5a users and typically it led to 10-20x decrease in write speed. PR: 13894 Submitted by: Maxim Sobolev ----- end log msg ----- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 8: 1:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from awfulhak.org (dynamic-74.max3-du-ws.dialnetwork.pavilion.co.uk [212.74.9.74]) by hub.freebsd.org (Postfix) with ESMTP id D891E1506C for ; Mon, 25 Oct 1999 08:01:22 -0700 (PDT) (envelope-from brian@Awfulhak.org) Received: from hak.lan.Awfulhak.org (root@hak.lan.Awfulhak.org [172.16.0.12]) by awfulhak.org (8.9.3/8.9.3) with ESMTP id QAA00333; Mon, 25 Oct 1999 16:01:20 +0100 (BST) (envelope-from brian@lan.awfulhak.org) Received: from hak.lan.Awfulhak.org (brian@localhost.lan.Awfulhak.org [127.0.0.1]) by hak.lan.Awfulhak.org (8.9.3/8.9.3) with ESMTP id HAA12679; Mon, 25 Oct 1999 07:51:30 -0700 (PDT) (envelope-from brian@hak.lan.Awfulhak.org) Message-Id: <199910251451.HAA12679@hak.lan.Awfulhak.org> X-Mailer: exmh version 2.1.0 09/18/1999 To: Jason Lewis Cc: Brian Somers , freebsd-net@FreeBSD.ORG, brian@hak.lan.Awfulhak.org Subject: Re: ppp In-Reply-To: Message from Jason Lewis of "Sun, 24 Oct 1999 18:07:19 MDT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 25 Oct 1999 07:51:30 -0700 From: Brian Somers Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'm afraid I haven't done any testing on a machine with 8Mb of RAM, but if you allocate a reasonable amount of swap there should be no problems (apart from speed problems). > Ok. I tried this and found out that my router is intermittant. I > suspect that I need more ram than 8 Mb. I could not find anything on how > much is needed with a router and ppp. I think I need atleast 32M for the > box to work with routing. Any insite on this? > > On Thu, 7 Oct 1999, Brian Somers wrote: > > > > How can one route public IP's over ppp without IP Masquerading? I tried > > > enabling forwarding, but it didnot work. Do I need to use pppd instead of > > > ppp? > > > > I'd advise using tcpdump (or enabling ppps tcp/ip logging) at each > > point of the packets journey to discover what's not forwarding it... > > > > Ppp is functionally the same as pppd in this respect. -- Brian Don't _EVER_ lose your sense of humour ! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 8:21:38 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpo.whistle.com (alpo.whistle.com [207.76.204.38]) by hub.freebsd.org (Postfix) with ESMTP id 6D3B414A1A; Mon, 25 Oct 1999 08:21:34 -0700 (PDT) (envelope-from julian@whistle.com) Received: from current1.whiste.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with ESMTP id IAA99744; Mon, 25 Oct 1999 08:18:28 -0700 (PDT) Date: Mon, 25 Oct 1999 08:18:28 -0700 (PDT) From: Julian Elischer To: "John E. Hein" Cc: Tony Jago , freebsd-net@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Re: Very Poor Samba -> Win9x performance [more] In-Reply-To: <14356.21920.670949.649870@Daffy.timing.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org his problem is in READ speed (read from disk, transmit on socket) I'm sure I have heard of this before.. On Mon, 25 Oct 1999, John E. Hein wrote: > Tony Jago wrote at 18:33 +1000 on Oct 25: > > Further to my problem with slow smbclient "put"'s. > > > > 1. FreeBSD 3.[23] machine -> Win98 slow (9k per second). > > 2. Solaris machine -> Win98 fast (500+k per second). > > 3. FreeBSD 2.2.7 machine -> Win98 fast (500+k per second). > > > > Just incase I was doing something insane, I installed a brand new > > 3.3-RELEASE machine running the GENERIC kernel, installed samba from the > > ports and did a smbclient "put" - still very slow performance. > > Make sure you've CVSup'd the latest samba port. On 9/24 patch-ah > was added which makes a mod to lib/util_sock.c to resolve this > problem on FreeBSD machines: > > Here's the log description of the patch. It seems to explain > exactly your problem, but you didn't give enough info about your > samba port installation to tell whether you got this patch or not. > > ----- start log msg for patch-ah ----- > revision 1.1 > date: 1999/09/24 04:29:42; author: cpiazza; state: Exp; > Add a patch to fix a problem with very low write speeds with > samba on freebsd. > > Following is description made by one of the Samba developers: > > "The problem is we switched to using recv() with the MSG_WAITALL flag > instead of read(). This makes Samba faster on most systems. On FreeBSD it > causes a massive slowdown and I don't know why. To fix it, change > the definition of MSG_WAITALL to be zero in lib/util_sock.c and > recompile 2.0.5a. I'm hoping someone from FreeBSD will get back to me > with some explanation. Regards, Jeremy Allison, Samba Team." > > This problem where observed by almost all samba-2.0.5a users and typically > it led to 10-20x decrease in write speed. > > PR: 13894 > Submitted by: Maxim Sobolev > ----- end log msg ----- > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 8:45:42 1999 Delivered-To: freebsd-net@freebsd.org Received: from pebkac.owp.csus.edu (pebkac.owp.csus.edu [130.86.232.245]) by hub.freebsd.org (Postfix) with ESMTP id 8F8BE150B4; Mon, 25 Oct 1999 08:45:37 -0700 (PDT) (envelope-from joseph.scott@owp.csus.edu) Received: from owp.csus.edu (mothra.ecs.csus.edu [130.86.76.220]) by pebkac.owp.csus.edu (8.9.3/8.9.3) with ESMTP id IAA52345; Mon, 25 Oct 1999 08:44:13 -0700 (PDT) Message-ID: <38147AC9.B644ECEB@owp.csus.edu> Date: Mon, 25 Oct 1999 15:44:09 +0000 From: Joseph Scott Organization: Water Programs - CSU Sacramento X-Mailer: Mozilla 4.61 [en] (X11; I; Linux 2.0.36 i386) X-Accept-Language: en MIME-Version: 1.0 To: Tony Jago Cc: Julian Elischer , freebsd-net@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Re: Very Poor Samba -> Win9x performance [more] References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Tony Jago wrote: > > Further to my problem with slow smbclient "put"'s. > > 1. FreeBSD 3.[23] machine -> Win98 slow (9k per second). > 2. Solaris machine -> Win98 fast (500+k per second). > 3. FreeBSD 2.2.7 machine -> Win98 fast (500+k per second). > > Just incase I was doing something insane, I installed a brand new > 3.3-RELEASE machine running the GENERIC kernel, installed samba from the > ports and did a smbclient "put" - still very slow performance. This is > very repeatable, I have tryed this on 3 different hardware platforms on > different networks and different network cards, they all do exactly the > same thing. > > Tweeking with the samba config file makes no difference, this must be > something way beyond that. We are talking 9k a second, not much more then > a modem out of 100M ethernet cards on switched networks! These machine > can ftp at rates closer to 5 megabytes a second and above. > > smbclient \\\\win98box\\test -N -c "put 1mfile" > > Could some reading this that has samba installed on a FreeBSD 3.3 machine > please try this just so that I know that I am not going insane? There was a patch added to the samba port ( after 3.3 I believe ) to address this very issue. I would try a fresh 3.3 install, cvsup the ports collection, install samba from the ports and try your test again. For info on the patch see : http://www.freebsd.org/cgi/cvsweb.cgi/ports/net/samba/patches/patch-ah I also believe this will be addressed in samba-2.0.6 I think. -- Joseph Scott joseph.scott@owp.csus.edu Office Of Water Programs - CSU Sacramento To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 8:55:33 1999 Delivered-To: freebsd-net@freebsd.org Received: from horst.bfd.com (horst.bfd.com [12.9.219.10]) by hub.freebsd.org (Postfix) with ESMTP id 3E5CC151DB; Mon, 25 Oct 1999 08:55:29 -0700 (PDT) (envelope-from ejs@bfd.com) Received: from HARLIE.bfd.com (bastion.bfd.com [12.9.219.14]) by horst.bfd.com (8.9.3/8.9.3) with ESMTP id IAA14375; Mon, 25 Oct 1999 08:54:07 -0700 (PDT) (envelope-from ejs@bfd.com) Date: Mon, 25 Oct 1999 08:54:07 -0700 (PDT) From: "Eric J. Schwertfeger" To: Tony Jago Cc: Julian Elischer , freebsd-net@FreeBSD.ORG, freebsd-questions@FreeBSD.ORG Subject: Re: Very Poor Samba -> Win9x performance [more] In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mon, 25 Oct 1999, Tony Jago wrote: > Just incase I was doing something insane, I installed a brand new > 3.3-RELEASE machine running the GENERIC kernel, installed samba from the > ports and did a smbclient "put" - still very slow performance. This is > very repeatable, I have tryed this on 3 different hardware platforms on > different networks and different network cards, they all do exactly the > same thing. > > Tweeking with the samba config file makes no difference, this must be > something way beyond that. We are talking 9k a second, not much more then > a modem out of 100M ethernet cards on switched networks! These machine > can ftp at rates closer to 5 megabytes a second and above. Someone stated that the Samba port had patches to fix the speed problem seen with FreeBSD (it's not just 3.3, it's 2.0.5a and any recent FreeBSD, AFAIK). What they failed to mention is that the Samba port was updated after 3.3-RELEASE, and it looks like said patches are part of the post-3.3R update. So try downloading the latest port, and installing from that. I'll be doing the same today. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 9:43:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from bomber.avantgo.com (ws1.avantgo.com [207.214.200.194]) by hub.freebsd.org (Postfix) with ESMTP id 6D2AB14D90 for ; Mon, 25 Oct 1999 09:43:19 -0700 (PDT) (envelope-from scott@avantgo.com) Received: from river ([10.0.128.30]) by bomber.avantgo.com (Netscape Messaging Server 3.5) with SMTP id 304 for ; Mon, 25 Oct 1999 09:37:50 -0700 Message-ID: <048d01bf1f07$e85749d0$1e80000a@avantgo.com> From: "Scott Hess" To: Subject: RFC 2140. Date: Mon, 25 Oct 1999 09:42:21 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Does FreeBSD, 3.3 or 4.0, contain anything like the optimizations discussed in RFC 2140? Basically, this RFC describes treating certain TCP control block information as host-pair shared rather than only used on a per-connection basis. Things like congestion control and suggested window size. I'm looking at it as a way to gain many of the positives of multiplexing things over long-running sockets, without having to build the application intelligence to multiplex over long-running sockets. Thanks, scott To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 9:55:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from fw.wintelcom.net (ns1.wintelcom.net [209.1.153.20]) by hub.freebsd.org (Postfix) with ESMTP id DEED4151DB for ; Mon, 25 Oct 1999 09:55:44 -0700 (PDT) (envelope-from bright@wintelcom.net) Received: from localhost (bright@localhost) by fw.wintelcom.net (8.9.3/8.9.3) with ESMTP id KAA03209; Mon, 25 Oct 1999 10:16:23 -0700 (PDT) Date: Mon, 25 Oct 1999 10:16:22 -0700 (PDT) From: Alfred Perlstein To: Tony Jago Cc: Julian Elischer , freebsd-net@FreeBSD.ORG Subject: Re: Very Poor Samba -> Win9x performance [more] In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mon, 25 Oct 1999, Tony Jago wrote: > > Further to my problem with slow smbclient "put"'s. > > 1. FreeBSD 3.[23] machine -> Win98 slow (9k per second). > 2. Solaris machine -> Win98 fast (500+k per second). > 3. FreeBSD 2.2.7 machine -> Win98 fast (500+k per second). > > Just incase I was doing something insane, I installed a brand new > 3.3-RELEASE machine running the GENERIC kernel, installed samba from the > ports and did a smbclient "put" - still very slow performance. This is > very repeatable, I have tryed this on 3 different hardware platforms on > different networks and different network cards, they all do exactly the > same thing. > > Tweeking with the samba config file makes no difference, this must be > something way beyond that. We are talking 9k a second, not much more then > a modem out of 100M ethernet cards on switched networks! These machine > can ftp at rates closer to 5 megabytes a second and above. > > smbclient \\\\win98box\\test -N -c "put 1mfile" > > Could some reading this that has samba installed on a FreeBSD 3.3 machine > please try this just so that I know that I am not going insane? After compiling the port please go into the work/samba*/lib dir, check the file "util_sock.c", make sure that around a line that looks like this: #endif /* WITH_SSL */ that below it there are 2 lines: #undef MSG_WAITALL #define MSG_WAITALL 0 (and please don't cross-post) -Alfred To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 9:56:11 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 52896150B9 for ; Mon, 25 Oct 1999 09:56:08 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id MAA79708; Mon, 25 Oct 1999 12:56:07 -0400 (EDT) (envelope-from wollman) Date: Mon, 25 Oct 1999 12:56:07 -0400 (EDT) From: Garrett Wollman Message-Id: <199910251656.MAA79708@khavrinen.lcs.mit.edu> To: "Scott Hess" Cc: Subject: RFC 2140. In-Reply-To: <048d01bf1f07$e85749d0$1e80000a@avantgo.com> References: <048d01bf1f07$e85749d0$1e80000a@avantgo.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Does FreeBSD, 3.3 or 4.0, contain anything like the optimizations discussed > in RFC 2140? No. Would you like to contribute code to do so? -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 10:18:45 1999 Delivered-To: freebsd-net@freebsd.org Received: from doggate.exchange.microsoft.com (doggate.exchange.microsoft.com [131.107.88.55]) by hub.freebsd.org (Postfix) with ESMTP id 4D04514BCD; Mon, 25 Oct 1999 10:18:22 -0700 (PDT) (envelope-from minwei@Exchange.Microsoft.com) Received: by doggate.exchange.microsoft.com with Internet Mail Service (5.5.2232.9) id <42Y7ABCC>; Mon, 25 Oct 1999 10:18:24 -0700 Message-ID: <01D6C7224936D211BA450000F805D5380F8D0486@TOTO> From: "Min Wei (Exchange)" To: freebsd-net@FreeBSD.ORG, "'hackers@freebsd.org'" Subject: FreeBSD reboots Date: Mon, 25 Oct 1999 10:18:13 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2232.9) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi. I have a situation where FreeBSD 3.2 reboots once a while (less than every 12 hours). My net environment is, FreeBSD 3.2 sits behind a Cisco LocalDirector. The FreeBSD machine is a Dell dual-proc with 512M RAM. I recompiled the kernel with SMP options and set MAXMEM to 512M (since by default FreeBSD thinks it only has 64M). I have apache installed as the Web server. However during our Web stress run, FreeBSD reboots periodically. I wonder if it's because a lot of TCP connections at FIN_WAIT_2 state, which causes the kernel crash. The load on the machine is not high (CPU is about 90% idle from top). I could only guess Cisco LD might not be configured properly which causes a lot of open connections. Anyone sees this kind of problem before? Any advice would be greatly appreciated. Thanks, --min To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 10:45: 0 1999 Delivered-To: freebsd-net@freebsd.org Received: from bomber.avantgo.com (ws1.avantgo.com [207.214.200.194]) by hub.freebsd.org (Postfix) with ESMTP id 15CCC15194 for ; Mon, 25 Oct 1999 10:44:54 -0700 (PDT) (envelope-from scott@avantgo.com) Received: from river ([10.0.128.30]) by bomber.avantgo.com (Netscape Messaging Server 3.5) with SMTP id 203; Mon, 25 Oct 1999 10:39:24 -0700 Message-ID: <051301bf1f10$82064510$1e80000a@avantgo.com> From: "Scott Hess" To: "Garrett Wollman" Cc: References: <048d01bf1f07$e85749d0$1e80000a@avantgo.com> <199910251656.MAA79708@khavrinen.lcs.mit.edu> Subject: Re: RFC 2140. Date: Mon, 25 Oct 1999 10:43:54 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'd love to, but my understanding of the kernel structures is not nearly deep enough, as of yet. Later, scott ----- Original Message ----- From: Garrett Wollman To: Scott Hess Cc: Sent: Monday, October 25, 1999 9:56 AM Subject: RFC 2140. > < said: > > > Does FreeBSD, 3.3 or 4.0, contain anything like the optimizations discussed > > in RFC 2140? > > No. Would you like to contribute code to do so? > > -GAWollman > > -- > Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same > wollman@lcs.mit.edu | O Siem / The fires of freedom > Opinions not those of| Dance in the burning flame > MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 10:52:34 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpo.whistle.com (alpo.whistle.com [207.76.204.38]) by hub.freebsd.org (Postfix) with ESMTP id E92A315194; Mon, 25 Oct 1999 10:52:26 -0700 (PDT) (envelope-from julian@whistle.com) Received: from current1.whiste.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with ESMTP id KAA10877; Mon, 25 Oct 1999 10:52:24 -0700 (PDT) Date: Mon, 25 Oct 1999 10:52:24 -0700 (PDT) From: Julian Elischer To: "Min Wei (Exchange)" Cc: freebsd-net@FreeBSD.ORG, "'hackers@freebsd.org'" Subject: Re: FreeBSD reboots In-Reply-To: <01D6C7224936D211BA450000F805D5380F8D0486@TOTO> Message-ID: MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="0-453316678-940873944=:7282" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. --0-453316678-940873944=:7282 Content-Type: TEXT/PLAIN; charset=US-ASCII why do you have a lot of connections in 'fin-wait-2 state? do you have windows clients? how many is 'a lot' if you leave the console on and so NOT have X11 runjing you may see mesages on the console. (possibly install DDB so that it doesn't fully reboot, but stops in the debugger) I have a patch to fix the fin-wait-2 problem.. I have included it. (It was last tested in 3.2) What it does is allow probing of the client so that the client will send an RST if the session is really closed, and if no response for a while, it will time out. You need to enable it in the config file too julian On Mon, 25 Oct 1999, Min Wei (Exchange) wrote: > Hi. I have a situation where FreeBSD 3.2 reboots once a while (less than > every > 12 hours). My net environment is, FreeBSD 3.2 sits behind a Cisco > LocalDirector. > The FreeBSD machine is a Dell dual-proc with 512M RAM. I recompiled the > kernel > with SMP options and set MAXMEM to 512M (since by default FreeBSD thinks it > only > has 64M). I have apache installed as the Web server. > > However during our Web stress run, FreeBSD reboots periodically. I wonder > if it's because a lot of TCP connections at FIN_WAIT_2 state, which causes > the kernel crash. The load on the machine is not high (CPU is about 90% idle > from > top). I could only guess Cisco LD might not be configured properly which > causes > a lot of open connections. > > Anyone sees this kind of problem before? Any advice would be greatly > appreciated. > > Thanks, > --min > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message > --0-453316678-940873944=:7282 Content-Type: TEXT/PLAIN; charset=US-ASCII; name="FINWAIT2.patch" Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename="FINWAIT2.patch" SW5kZXg6IG5ldGluZXQvdGNwX2lucHV0LmMNCj09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT0NClJDUyBmaWxlOiAvaG9tZS9uY3ZzL3NyYy9zeXMvbmV0aW5ldC90 Y3BfaW5wdXQuYyx2DQpyZXRyaWV2aW5nIHJldmlzaW9uIDEuODQNCmRpZmYg LWMgLXIxLjg0IG5ldGluZXQvdGNwX2lucHV0LmMNCioqKiBuZXRpbmV0L3Rj cF9pbnB1dC5jCTE5OTkvMDIvMDYgMDA6NDc6NDUJMS44NA0KLS0tIG5ldGlu ZXQvdGNwX2lucHV0LmMJMTk5OS8wNC8wOCAwMDo0NTo1NA0KKioqKioqKioq KioqKioqDQoqKiogODgsOTMgKioqKg0KLS0tIDg4LDk5IC0tLS0NCiAgU1lT Q1RMX0lOVChfbmV0X2luZXRfdGNwLCBPSURfQVVUTywgZGVsYXllZF9hY2ss IENUTEZMQUdfUlcsIA0KICAJJnRjcF9kZWxhY2tfZW5hYmxlZCwgMCwgIiIp Ow0KICANCisgI2lmZGVmCVRDUF9VU09GVF9CVUcNCisgaW50IHRjcF9maW5f cmV0cnlfZW5hYmxlID0gMDsNCisgU1lTQ1RMX0lOVChfbmV0X2luZXRfdGNw LCBPSURfQVVUTywgZmluX3JldHJ5LCBDVExGTEFHX1JXLCANCisgCSZ0Y3Bf ZmluX3JldHJ5X2VuYWJsZSwgMCwgIiIpOw0KKyAjZW5kaWYJLyogVENQX1VT T0ZUX0JVRyAqLw0KKyANCiAgdV9sb25nCXRjcF9ub3c7DQogIHN0cnVjdCBp bnBjYmhlYWQgdGNiOw0KICBzdHJ1Y3QgaW5wY2JpbmZvIHRjYmluZm87DQoq KioqKioqKioqKioqKioNCioqKiAxNDk1LDE1MDQgKioqKg0KLS0tIDE1MDEs MTU0MCAtLS0tDQogIAkJCQkgKiBzcGVjaWZpY2F0aW9uLCBidXQgaWYgd2Ug ZG9uJ3QgZ2V0IGEgRklODQogIAkJCQkgKiB3ZSdsbCBoYW5nIGZvcmV2ZXIu DQogIAkJCQkgKi8NCisgI2lmZGVmIFRDUF9VU09GVF9CVUcNCisgCQkJCS8q DQorIAkJCQkgKiBXYWl0IGxvbmdlciBhbmQgbG9uZ2VyIGZvciB0aGUgb3Ro ZXINCisgCQkJCSAqIGVuZCB0byByZXNwb25kIHdpdGggc29tZXRoaW5nLg0K KyAJCQkJICogRXZlbnR1YWxseSB0aGV5IHNob3VsZCBlaXRoZXINCisgCQkJ CSAqIFJTVCBvciBGSU4uIElmIHRoZXkgYXJlIHN0aWxsIGFsaXZlDQorIAkJ CQkgKiBhbmQgYWN0dWFsbHkgd2FudCB1cyB0byByZW1haW4gaW4gdGhpcw0K KyAJCQkJICogc3RhdGUsIHRoZXkgd2lsbCBrZWVwIEFDS2luZyBhbmQNCisg CQkJCSAqIHdlJ2xsIHN0YXkgaGVyZSBpbmRlZmluaXRseS4NCisgCQkJCSAq IElmIHRoZXkgZG9uJ3QgcmVzcG9uZCBhdCBhbGwsIHdlIHdpbGwNCisgCQkJ CSAqIHJldmVydCB0byBGSU5fV0FJVF8xIGFuZCBldmVudHVhbGx5DQorIAkJ CQkgKiB0aW1lIG91dCBhcyBpdCB3b3VsZC4gSG93IHRvIGNvcGUgd2l0aA0K KyAJCQkJICogdGhlIGNhc2Ugb2YgYnJva2VuIGNsaWVudHMgd2hvIGFyZSBz dGlsbA0KKyAJCQkJICogYWxpdmUgYnV0IG5ldmVyIEZJTiBpcyBhcmd1YWJs ZS4gQ2VydGFpbmx5DQorIAkJCQkgKiBpZiB3ZSd2ZSBjbG9zZWQgb3VyIGVu ZCBlbnRpcmVseSwgd2UNCisgCQkJCSAqIG1pZ2h0IGFzIHdlbGwganVzdCBj bG9zZSB0aGUgY29ubmVjdGlvbi4NCisgCQkJCSAqLw0KKyAJCQkJaWYgKHRj cF9maW5fcmV0cnlfZW5hYmxlKSB7DQorIAkJCQl0cC0+dF90aW1lcltUQ1BU XzJNU0xdID0gDQorIAkJCQkJKCh0cC0+dF9pZGxlID4gVENQVFZfTVNMKSA/ DQorIAkJCQkJdHAtPnRfaWRsZSA6IFRDUFRWX01TTCk7DQorIAkJCQl9IGVs c2Ugew0KKyAJCQkJCWlmIChzby0+c29fc3RhdGUgJiBTU19DQU5UUkNWTU9S RSkgew0KKyAJCQkJCQlzb2lzZGlzY29ubmVjdGVkKHNvKTsNCisgCQkJCQkJ dHAtPnRfdGltZXJbVENQVF8yTVNMXQ0KKyAJCQkJCQkJCT0gdGNwX21heGlk bGU7DQorIAkJCQkJfQ0KKyAJCQkJfQ0KKyAjZWxzZQkvKiBUQ1BfVVNPRlRf QlVHICovDQogIAkJCQlpZiAoc28tPnNvX3N0YXRlICYgU1NfQ0FOVFJDVk1P UkUpIHsNCiAgCQkJCQlzb2lzZGlzY29ubmVjdGVkKHNvKTsNCiAgCQkJCQl0 cC0+dF90aW1lcltUQ1BUXzJNU0xdID0gdGNwX21heGlkbGU7DQogIAkJCQl9 DQorICNlbmRpZgkvKiBUQ1BfVVNPRlRfQlVHICovDQogIAkJCQl0cC0+dF9z dGF0ZSA9IFRDUFNfRklOX1dBSVRfMjsNCiAgCQkJfQ0KICAJCQlicmVhazsN CkluZGV4OiBuZXRpbmV0L3RjcF90aW1lci5jDQo9PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09DQpSQ1MgZmlsZTogL2hvbWUvbmN2cy9zcmMvc3lzL25ldGluZXQv dGNwX3RpbWVyLmMsdg0KcmV0cmlldmluZyByZXZpc2lvbiAxLjI4DQpkaWZm IC1jIC1yMS4yOCBuZXRpbmV0L3RjcF90aW1lci5jDQoqKiogbmV0aW5ldC90 Y3BfdGltZXIuYwkxOTk4LzA0LzI0IDA5OjI1OjM1CTEuMjgNCi0tLSBuZXRp bmV0L3RjcF90aW1lci5jCTE5OTkvMDQvMDggMDA6NDU6NTQNCioqKioqKioq KioqKioqKg0KKioqIDIxMywyMjIgKioqKg0KICAJICogY29udHJvbCBibG9j ay4gIE90aGVyd2lzZSwgY2hlY2sgYWdhaW4gaW4gYSBiaXQuDQogIAkgKi8N CiAgCWNhc2UgVENQVF8yTVNMOg0KISAJCWlmICh0cC0+dF9zdGF0ZSAhPSBU Q1BTX1RJTUVfV0FJVCAmJg0KISAJCSAgICB0cC0+dF9pZGxlIDw9IHRjcF9t YXhpZGxlKQ0KISAJCQl0cC0+dF90aW1lcltUQ1BUXzJNU0xdID0gdGNwX2tl ZXBpbnR2bDsNCiEgCQllbHNlDQogIAkJCXRwID0gdGNwX2Nsb3NlKHRwKTsN CiAgCQlicmVhazsNCiAgDQotLS0gMjEzLDI0NiAtLS0tDQogIAkgKiBjb250 cm9sIGJsb2NrLiAgT3RoZXJ3aXNlLCBjaGVjayBhZ2FpbiBpbiBhIGJpdC4N CiAgCSAqLw0KICAJY2FzZSBUQ1BUXzJNU0w6DQohIAkJaWYgKHRwLT50X3N0 YXRlICE9IFRDUFNfVElNRV9XQUlUDQohIAkJJiYgdHAtPnRfaWRsZSA8PSB0 Y3BfbWF4aWRsZSkgew0KISAjaWZkZWYgVENQX1VTT0ZUX0JVRw0KISAJCQlp ZiAodGNwX2Zpbl9yZXRyeV9lbmFibGUNCiEgCQkJJiYgKHRwLT50X3N0YXRl ID09IFRDUFNfRklOX1dBSVRfMikpIHsNCiEgCQkJCS8qDQohIAkJCQkgKiBX ZSd2ZSB0aW1lZCBvdXQgd2FpdGluZyBmb3IgdGhlIG90aGVyIGVuZCANCiEg CQkJCSAqIHRvIGZpbmlzaCB1cC4gUXVpdGUgcG9zc2libHkgaXQncyBhIFdp bjl4DQohIAkJCQkgKiBtYWNoaW5lLg0KISAJCQkJICogSWYgc28gd2UgY291 bGQgYmUgd2FpdGluZyBoZXJlIGZvcmV2ZXIuDQohIAkJCQkgKiBQcmV0ZW5k IHdlIHdlcmUgbmV2ZXIgYWNrJ2QgYW5kIHJlc2V0DQohIAkJCQkgKiBvdXJz ZWx2ZXMgdG8gYSByZXRyeSBvZiBGSU5fV0FJVF8xLiBJZg0KISAJCQkJICog aXQncyBzdGlsbCBhbGl2ZSwgdGhpcyBzaG91bGQgYXQgbGVhc3QNCiEgCQkJ CSAqIGVsaWNpdCBhIFJTVCBmcm9tIGl0IHdoaWNoDQohIAkJCQkgKiB3aWxs IGxldCB1cyBrbm93IHdlIGNhbiBzaHV0IGRvd24uDQohIAkJCQkgKiBJZiBp dCBoYXMgb25seSBkb25lIGEgaGFsZiBjbG9zZSwNCiEgCQkJCSAqIGl0J2xs IEFDSyBvdXIgcmV0cmllcyBzbyB3ZSdsbA0KISAJCQkJICoga2VlcCB3YWl0 aW5nIGluIEZJTl9XQUlUXzIuDQohIAkJCQkgKiBJZiBpdCdzIGRlYWQsIHdl J2xsIHRpbWUgb3V0Lg0KISAJCQkJICovDQohIAkJCQl0cC0+dF9zdGF0ZSA9 IFRDUFNfRklOX1dBSVRfMTsNCiEgCQkJCXRwLT50X2ZsYWdzICY9IH5URl9T RU5URklOOw0KISAJCQkJdHAtPnNuZF91bmEgPSAodHAtPnNuZF9ueHQgLT0g MSk7DQohIAkJCQl0Y3Bfb3V0cHV0KHRwKTsNCiEgCQkJfSBlbHNlDQohICNl bmRpZg0KISAJCQkJdHAtPnRfdGltZXJbVENQVF8yTVNMXSA9IHRjcF9rZWVw aW50dmw7DQohIAkJfSBlbHNlDQogIAkJCXRwID0gdGNwX2Nsb3NlKHRwKTsN CiAgCQlicmVhazsNCiAgDQpJbmRleDogbmV0aW5ldC90Y3BfdmFyLmgNCj09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT0NClJDUyBmaWxlOiAvaG9tZS9uY3ZzL3Ny Yy9zeXMvbmV0aW5ldC90Y3BfdmFyLmgsdg0KcmV0cmlldmluZyByZXZpc2lv biAxLjUwDQpkaWZmIC1jIC1yMS41MCBuZXRpbmV0L3RjcF92YXIuaA0KKioq IG5ldGluZXQvdGNwX3Zhci5oCTE5OTkvMDIvMTYgMTA6NDk6NTIJMS41MA0K LS0tIG5ldGluZXQvdGNwX3Zhci5oCTE5OTkvMDQvMDggMDA6NDU6NTQNCioq KioqKioqKioqKioqKg0KKioqIDMzNCwzMzkgKioqKg0KLS0tIDMzNCwzNDAg LS0tLQ0KICBleHRlcm4JaW50IHRjcF9tc3NkZmx0OwkvKiBYWFggKi8NCiAg ZXh0ZXJuCXVfbG9uZyB0Y3Bfbm93OwkJLyogZm9yIFJGQyAxMzIzIHRpbWVz dGFtcHMgKi8NCiAgZXh0ZXJuCWludCB0Y3BfZGVsYWNrX2VuYWJsZWQ7DQor IGV4dGVybglpbnQgdGNwX2Zpbl9yZXRyeV9lbmFibGU7DQogIA0KICB2b2lk CSB0Y3BfY2FuY2VsdGltZXJzIF9fUCgoc3RydWN0IHRjcGNiICopKTsNCiAg c3RydWN0IHRjcGNiICoNCg== --0-453316678-940873944=:7282-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Oct 25 22:46: 9 1999 Delivered-To: freebsd-net@freebsd.org Received: from jason.argos.org (a13b146.neo.rr.com [204.210.197.146]) by hub.freebsd.org (Postfix) with ESMTP id 9399214CDF for ; Mon, 25 Oct 1999 22:46:06 -0700 (PDT) (envelope-from mike@argos.org) Received: from localhost (mike@localhost) by jason.argos.org (8.9.1/8.9.1) with ESMTP id BAA30914; Tue, 26 Oct 1999 01:45:42 -0400 Date: Tue, 26 Oct 1999 01:45:42 -0400 (EDT) From: Mike Nowlin To: Jason Lewis Cc: Brian Somers , freebsd-net@FreeBSD.ORG Subject: Re: ppp In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Ok. I tried this and found out that my router is intermittant. I > suspect that I need more ram than 8 Mb. I could not find anything on how > much is needed with a router and ppp. I think I need atleast 32M for the > box to work with routing. Any insite on this? You should be able to do this with 16, and probably 8.... The stuff simply isn't that big, especially if you strip it (dunno if it does this by default.) If I can get diskless X terminals running in 16 megs, a router shouldn't be that hard - make sure your kernel doesn't have extra junk in it, and kill any programs that you don't need... mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 0:25:55 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id AAB4D14CAC for ; Tue, 26 Oct 1999 00:25:53 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id AAA18890 for ; Tue, 26 Oct 1999 00:25:53 -0700 (PDT) Message-ID: <1df701bf1f84$32751020$0201010a@fuzzer.com> From: "Craig Critchley" To: Subject: FTP Net Performance Date: Tue, 26 Oct 1999 00:32:02 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I've got one FreeBSD 3.2-RELEASE machine, an NT 4 machine, and a Win98 machine. The BSD serves as a web/ftp/mail server. All have NetGear 100BaseT cards and are connected through a DLink 10/100Mb hub ("with switch" e.g. one of those that makes a 100mb and 10mb segments and switches between them). Running an FTP client on either the Win98 machine or the NT machine, sending files *to* the BSD machine goes full speed. However, getting files *from* the BSD ftpd runs at less than 10 kbytes/sec. Monstrous! (And the blinky lights on the hub are barely going.) The web server seems to have the same problem, e.g. if I put the same files in a www directory and try to download them with the browser things go about the same painful speed. Transfers between Windows machines seem normal (even copies through the 10mb segment are a lot faster). Any suggestions profoundly appreciated. ...Craig To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 1:11:57 1999 Delivered-To: freebsd-net@freebsd.org Received: from acacia.cts.ucla.edu (acacia.ctslab.ucla.edu [164.67.61.44]) by hub.freebsd.org (Postfix) with ESMTP id 90FC014A10 for ; Tue, 26 Oct 1999 01:11:53 -0700 (PDT) (envelope-from denis@acacia.cts.ucla.edu) Received: from localhost (denis@localhost) by acacia.cts.ucla.edu (8.9.3/8.8.8) with ESMTP id BAA39208; Tue, 26 Oct 1999 01:13:53 -0700 (PDT) (envelope-from denis@acacia.cts.ucla.edu) Date: Tue, 26 Oct 1999 01:13:52 -0700 (PDT) From: Denis DeLaRoca To: spork Cc: freebsd-net@FreeBSD.ORG Subject: Re: PPPoE In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, 1 Oct 1999, spork wrote: > It seems more and more ADSL providers in the US are moving from bridged > IP over ethernet to PPP over Ethernet as they dump whatever clunky > solutions they started with and move to the RedBack "subscriber management > system". The idea it seems is to simulate the familiar dialup connection. > This lets you hand out dynamic addresses, dump idle users, discourage > servers, track usage, hamper NAT, and (the relevant part) discourage > people from connecting with anything but "supported" OS's. How does it hamper NAT, could you elaborate? -- Denis To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 1:45: 9 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpo.whistle.com (alpo.whistle.com [207.76.204.38]) by hub.freebsd.org (Postfix) with ESMTP id 704AF15348 for ; Tue, 26 Oct 1999 01:45:00 -0700 (PDT) (envelope-from julian@whistle.com) Received: from home.elischer.org (home.elischer.org [207.76.204.203]) by alpo.whistle.com (8.9.1a/8.9.1) with ESMTP id BAA48113; Tue, 26 Oct 1999 01:44:58 -0700 (PDT) Date: Tue, 26 Oct 1999 01:44:59 -0700 (PDT) From: Julian Elischer X-Sender: julian@home.elischer.org To: Denis DeLaRoca Cc: net@freebsd.org Subject: Re: PPPOE report. In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org yes there is news.. The PPPOE node negotiated a session on the test setup and I'm just cleaning up some stuff. Brian is doing some work on the ppp code to support it and when I'm finished doing this I'll send him some support code. On Tue, 26 Oct 1999, Denis DeLaRoca wrote: > > On Sat, 16 Oct 1999, Julian Elischer wrote: > > > > > I am now at the stage of testing, I have an offer of a test machine with > > both ISDN, and DSL with PPPOE which I will be using to test. > > > > reports in a day or so hopefully. > > Any news yet? > > How do you envision that doing a FreeBSD net-install would work over a > PPPoE aDSL connection? You'd have to have support PPPoE on the install > floppy, no? Unfortunatly yes, but that should be possible. The ppp will negotiate an IP address so it should work. > > -- Denis > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 1:49:43 1999 Delivered-To: freebsd-net@freebsd.org Received: from acacia.cts.ucla.edu (acacia.ctslab.ucla.edu [164.67.61.44]) by hub.freebsd.org (Postfix) with ESMTP id 0D48915348 for ; Tue, 26 Oct 1999 01:49:38 -0700 (PDT) (envelope-from denis@acacia.cts.ucla.edu) Received: from localhost (denis@localhost) by acacia.cts.ucla.edu (8.9.3/8.8.8) with ESMTP id BAA39277; Tue, 26 Oct 1999 01:51:39 -0700 (PDT) (envelope-from denis@acacia.cts.ucla.edu) Date: Tue, 26 Oct 1999 01:51:39 -0700 (PDT) From: Denis DeLaRoca To: Julian Elischer Cc: net@freebsd.org Subject: Re: PPPOE report. In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, 26 Oct 1999, Julian Elischer wrote: > yes there is news.. > > The PPPOE node negotiated a session on the test setup and I'm just > cleaning up some stuff. Great! Just now I stumbled upon the Zyxel web site (www.zyxel.com) and their Prestige 641 aDSL router claims to be the first/only router supporting PPPoE. -- Denis To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 1:54:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpo.whistle.com (alpo.whistle.com [207.76.204.38]) by hub.freebsd.org (Postfix) with ESMTP id 0E2AF15348 for ; Tue, 26 Oct 1999 01:54:42 -0700 (PDT) (envelope-from julian@whistle.com) Received: from home.elischer.org (home.elischer.org [207.76.204.203]) by alpo.whistle.com (8.9.1a/8.9.1) with ESMTP id BAA48294; Tue, 26 Oct 1999 01:54:29 -0700 (PDT) Date: Tue, 26 Oct 1999 01:54:30 -0700 (PDT) From: Julian Elischer X-Sender: julian@home.elischer.org To: Denis DeLaRoca Cc: net@FreeBSD.ORG Subject: Re: PPPOE report. In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, 26 Oct 1999, Denis DeLaRoca wrote: > > On Tue, 26 Oct 1999, Julian Elischer wrote: > > > yes there is news.. > > > > The PPPOE node negotiated a session on the test setup and I'm just > > cleaning up some stuff. > > Great! > > Just now I stumbled upon the Zyxel web site (www.zyxel.com) and their > Prestige 641 aDSL router claims to be the first/only router supporting > PPPoE. Not quite.. the telco's must be supporting it as they are doing it.. > > -- Denis > > > > > > 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 From owner-freebsd-net Tue Oct 26 7:19:33 1999 Delivered-To: freebsd-net@freebsd.org Received: from etek.chalmers.se (quarl0.etek.chalmers.se [129.16.32.20]) by hub.freebsd.org (Postfix) with ESMTP id 2A48914E4C for ; Tue, 26 Oct 1999 07:19:25 -0700 (PDT) (envelope-from em8david@etek.chalmers.se) Received: from quarl0.etek.chalmers.se (_7-C3C0@quarl0 [129.16.32.20]) by etek.chalmers.se (8.8.8/8.8.8) with ESMTP id QAA28633 for ; Tue, 26 Oct 1999 16:19:24 +0200 (MET DST) Received: from localhost (em8david@localhost) by quarl0.etek.chalmers.se (8.8.8/8.8.8) with SMTP id QAA28637 for ; Tue, 26 Oct 1999 16:19:24 +0200 (MET DST) X-Authentication-Warning: quarl0.etek.chalmers.se: em8david owned process doing -bs Date: Tue, 26 Oct 1999 16:19:23 +0200 (MET DST) From: David Przybylak X-Sender: em8david@quarl0 To: freebsd-net@FreeBSD.ORG Subject: Window scale option Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I am working on a Long Fat Network and I would like to be able to advertise large windows thanks to the window scale option of TCP. I have increased both the receiving and the sending buffer size to 200000 bytes with the commands "sysctl -w net.inet.tcp.sendspace=200000" and "sysctl -w net.inet.tcp.sendspace=200000" on my two computers (FreeBSD 3.2). I am using the "sock" program to create a socket between the two computers and both SO_SNDBUF and SO_RCVBUF are set to 200000 on the two computers while creating the socket (I use the options -R200000 and -S200000 of "sock" to do that). I thought that TCP automatically set the scale factor based on the size of the receive buffer. So, I expected TCP to advertise a window of 65535*2^2 and to see "wscale 2" in the option field of the SYN segment. But the problem is that, looking at the tcpdump output, it seems that TCP doesn't send any scale factor with the SYN, and the window advertised is only 65535 bytes long. Does anyone understand what the problem is? How can I use the window scale option? How can I send window advertisements up to 264144 bytes (which is the maximum value fixed by the sysctl MIB variable "kern.ipc.maxsockbuf")? Regards, David Przybylak To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 7:53:56 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 1D36514A2C for ; Tue, 26 Oct 1999 07:53:49 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id KAA82743; Tue, 26 Oct 1999 10:53:47 -0400 (EDT) (envelope-from wollman) Date: Tue, 26 Oct 1999 10:53:47 -0400 (EDT) From: Garrett Wollman Message-Id: <199910261453.KAA82743@khavrinen.lcs.mit.edu> To: David Przybylak Cc: freebsd-net@FreeBSD.ORG Subject: Window scale option In-Reply-To: References: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Does anyone understand what the problem is? How can I use the window > scale option? How can I send window advertisements up to 264144 bytes > (which is the maximum value fixed by the sysctl MIB variable > "kern.ipc.maxsockbuf")? You can't actually set them that high using the default maxsockbuf because TCP will round the window up to a whole number of segments. On my 4.0-current system: root@khavrinen(1)# sysctl -w net.inet.tcp.rfc1323=1 net.inet.tcp.rfc1323: 0 -> 1 root@khavrinen(4)# sysctl -w kern.ipc.maxsockbuf=524288 kern.ipc.maxsockbuf: 262144 -> 524288 root@khavrinen(5)# tcpdump -i lo0 -w /tmp/foo -s 128 tcp port 5001 tcpdump: listening on lo0 ^C 2074 packets received by filter 0 packets dropped by kernel root@khavrinen(6)# tcpdump -r /tmp/foo | more 10:38:06.070868 localhost.lcs.mit.edu.udt_os > localhost.lcs.mit.edu.commplex-link: S 228218603:228218603(0) win 16384 (DF) 10:38:06.070993 localhost.lcs.mit.edu.commplex-link > localhost.lcs.mit.edu.udt_os: S 228306224:228306224(0) ack 228218604 win 57344 (DF) On the receiving side: wollman@khavrinen(3)$ ttcp -b 262144 -r -s -v -f m ttcp-r: buflen=8192, nbuf=2048, align=16384/0, port=5001, sockbufsize=262144 tcp ttcp-r: socket ttcp-r: rcvbuf ttcp-r: accept from 127.0.0.1 ttcp-r: 16777216 bytes in 0.52 real seconds = 246.88 Mbit/sec +++ ttcp-r: 16777216 bytes in 0.22 CPU seconds = 586.21 Mbit/cpu sec ttcp-r: 2325 I/O calls, msec/call = 0.23, calls/sec = 4484.35 ttcp-r: 0.0user 0.2sys 0:00real 41% 128i+228d 422maxrss 0+2pf 553+236csw ttcp-r: buffer address 0x28000 On the sending side: wollman@khavrinen(2)$ ttcp -t -s -v -f m -b 262144 localhost ttcp-t: buflen=8192, nbuf=2048, align=16384/0, port=5001, sockbufsize=262144 tcp -> localhost ttcp-t: socket ttcp-t: sndbuf ttcp-t: connect ttcp-t: 16777216 bytes in 0.50 real seconds = 254.00 Mbit/sec +++ ttcp-t: 16777216 bytes in 0.22 CPU seconds = 568.96 Mbit/cpu sec ttcp-t: 2048 I/O calls, msec/call = 0.25, calls/sec = 4064.04 ttcp-t: 0.0user 0.2sys 0:00real 42% 181i+330d 86maxrss 0+2pf 232+557csw ttcp-t: buffer address 0x2c000 (This is on a three-year-old Pentium Pro on an Orion motherboard. A convenient 400-MHz Celeron-A gave about twice the speed.) -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 9: 6:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id A3B7C14BED for ; Tue, 26 Oct 1999 09:06:38 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id MAA83232; Tue, 26 Oct 1999 12:06:36 -0400 (EDT) (envelope-from wollman) Date: Tue, 26 Oct 1999 12:06:36 -0400 (EDT) From: Garrett Wollman Message-Id: <199910261606.MAA83232@khavrinen.lcs.mit.edu> To: "Craig Critchley" Cc: Subject: FTP Net Performance In-Reply-To: <1df701bf1f84$32751020$0201010a@fuzzer.com> References: <1df701bf1f84$32751020$0201010a@fuzzer.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Any suggestions profoundly appreciated. Whenever TCP behaves strangely, your first step should be to capture a trace with tcpdump(8) and use tcpdump2xplot.pl and xplot(1) (part of /usr/ports/math/xplot) to see what's going on. The visual analysis technique is developed in Tim Shepard's S.M. thesis, references to which are available in the xplot tarball. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 14: 5:18 1999 Delivered-To: freebsd-net@freebsd.org Received: from proxy4.ba.best.com (proxy4.ba.best.com [206.184.139.15]) by hub.freebsd.org (Postfix) with ESMTP id 7B28D14FE7; Tue, 26 Oct 1999 14:05:04 -0700 (PDT) (envelope-from ssamalin@ionet.net) Received: from ionet.net (sam.ops.best.com [205.149.163.53]) by proxy4.ba.best.com (8.9.3/8.9.2/best.out) with ESMTP id OAA13326; Tue, 26 Oct 1999 14:02:51 -0700 (PDT) Message-ID: <381616CE.A9C79619@ionet.net> Date: Tue, 26 Oct 1999 17:02:06 -0400 From: Sam Samalin X-Mailer: Mozilla 4.6 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-security@freebsd.org, freebsd-net@freebsd.org Subject: ftp with ipfw fwd Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org My ftp server isn't using the "PORT" port to after the PORT call, it uses 20: "Can't create data socket (n.n.n.n,20) : Can't assign requested address." I'm using ipfw fwd. Do I need a rule? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 16:57:45 1999 Delivered-To: freebsd-net@freebsd.org Received: from pail.ircache.net (pail.scd.ucar.edu [128.117.28.5]) by hub.freebsd.org (Postfix) with ESMTP id E61DC14FCB for ; Tue, 26 Oct 1999 16:57:43 -0700 (PDT) (envelope-from rousskov@ircache.net) Received: from localhost (rousskov@localhost) by pail.ircache.net (8.9.2/8.8.7) with ESMTP id RAA48753 for ; Tue, 26 Oct 1999 17:57:43 -0600 (MDT) (envelope-from rousskov@ircache.net) Date: Tue, 26 Oct 1999 17:57:43 -0600 (MDT) From: Alex Rousskov To: freebsd-net@freebsd.org Subject: TCP throughput vs number of aliases Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi there, We use FreeBSD 3.1 to benchmark Web caching proxies. We want to use a [very] large number of IP addresses for simulated clients and servers in the benchmark setup. However, we are limited by the number of machines we can dedicate to the tests. Naturally, we are using interface aliases to create IP addresses. With the number of aliases above 600+ we are monitoring a sudden and significant drop in TCP throughput (measured using simple netperf tests). The graph and some details of the setup are at http://www.ircache.net/~wessels/Junk/aliases/ We would like to use more than 600 aliases; probably 2,000-5,000 aliases per box without the loss of performance. Can anybody recommend a solution? Thank you, Alex. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 16:58: 1 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id 239221500C for ; Tue, 26 Oct 1999 16:57:56 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id QAA09715; Tue, 26 Oct 1999 16:57:54 -0700 (PDT) Message-ID: <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> From: "Craig Critchley" To: "Garrett Wollman" , References: <1df701bf1f84$32751020$0201010a@fuzzer.com> <199910261606.MAA83232@khavrinen.lcs.mit.edu> Subject: Re: FTP Net Performance Date: Tue, 26 Oct 1999 17:04:05 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Thanks. I guess I need to track down the thesis and give it a read - the graphs themselves are relatively uninteresting - flat lines at large negative numbers. (This is a small enough network that I can make sure there is more or less nothing else happening when I try to FTP something.) I got conflicting suggestions from a couple people about whether to use full or half-duplex. The device defaults to half-duplexing, which seems to be correct for a hub (the two machines do not have the switch between them as they are both on the 100Mb segment). I'm not too concerned about 50Mb vs. 100Mb, or even 10Mb vs. 100Mb - I'm getting barely 100kilobits. Three orders of magnitude difference. But uploading to the BSD machine goes full speed. (For the record, ifconfig and the hub blinkies agree that I have a 100Mb link.) Looking at netstat -i, I see that I'm getting Oerrs of about 15% of Opkts, but no Ierrs. There is an inconsequential number of collisions. Perhaps the error rate is to blame for the problem. What sorts of things do these errors represent? Thanks, ...Craig ----- Original Message ----- From: Garrett Wollman To: Craig Critchley Cc: Sent: Tuesday, October 26, 1999 9:06 AM Subject: FTP Net Performance > < said: > > > Any suggestions profoundly appreciated. > > Whenever TCP behaves strangely, your first step should be to capture a > trace with tcpdump(8) and use tcpdump2xplot.pl and xplot(1) (part of > /usr/ports/math/xplot) to see what's going on. The visual analysis > technique is developed in Tim Shepard's S.M. thesis, references to > which are available in the xplot tarball. > > -GAWollman > > -- > Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same > wollman@lcs.mit.edu | O Siem / The fires of freedom > Opinions not those of| Dance in the burning flame > MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick > > > 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 From owner-freebsd-net Tue Oct 26 17:25:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from implode.root.com (root.com [209.102.106.178]) by hub.freebsd.org (Postfix) with ESMTP id 3229214C92 for ; Tue, 26 Oct 1999 17:25:22 -0700 (PDT) (envelope-from dg@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.8/8.8.5) with ESMTP id RAA23991; Tue, 26 Oct 1999 17:21:54 -0700 (PDT) Message-Id: <199910270021.RAA23991@implode.root.com> To: "Craig Critchley" Cc: "Garrett Wollman" , net@FreeBSD.ORG Subject: Re: FTP Net Performance In-reply-to: Your message of "Tue, 26 Oct 1999 17:04:05 PDT." <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> From: David Greenman Reply-To: dg@root.com Date: Tue, 26 Oct 1999 17:21:54 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >Thanks. I guess I need to track down the thesis and give it a read - the >graphs themselves are relatively uninteresting - flat lines at large >negative numbers. (This is a small enough network that I can make sure >there is more or less nothing else happening when I try to FTP something.) > >I got conflicting suggestions from a couple people about whether to use full >or half-duplex. The device defaults to half-duplexing, which seems to be >correct for a hub (the two machines do not have the switch between them as >they are both on the 100Mb segment). I'm not too concerned about 50Mb vs. >100Mb, or even 10Mb vs. 100Mb - I'm getting barely 100kilobits. Three >orders of magnitude difference. But uploading to the BSD machine goes full >speed. (For the record, ifconfig and the hub blinkies agree that I have a >100Mb link.) > >Looking at netstat -i, I see that I'm getting Oerrs of about 15% of Opkts, >but no Ierrs. There is an inconsequential number of collisions. Perhaps >the error rate is to blame for the problem. What sorts of things do these >errors represent? You shouldn't be getting any output errors - that is definately indicating a problem. Have you tried replacing the cable? What sort of motherboard is the card plugged into? -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project - http://www.freebsd.org Creator of high-performance Internet servers - http://www.terasolutions.com Pave the road of life with opportunities. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 17:35:28 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail-relay2.yahoo.com (mail-relay2.yahoo.com [206.251.17.77]) by hub.freebsd.org (Postfix) with ESMTP id A960514C92 for ; Tue, 26 Oct 1999 17:35:25 -0700 (PDT) (envelope-from jayanth@yahoo-inc.com) Received: from borogove.yahoo.com (borogove.yahoo.com [205.216.162.65]) by mail-relay2.yahoo.com (8.9.3/8.9.3) with ESMTP id RAA06352; Tue, 26 Oct 1999 17:35:25 -0700 (PDT) Received: from yahoo-inc.com (milk.yahoo.com [206.132.89.117]) by borogove.yahoo.com (8.9.3/8.9.3) with ESMTP id RAA21848; Tue, 26 Oct 1999 17:35:25 -0700 (PDT) Message-ID: <381648CC.B57E06BF@yahoo-inc.com> Date: Tue, 26 Oct 1999 17:35:24 -0700 From: jayanth X-Mailer: Mozilla 4.08 [en] (X11; I; FreeBSD 2.2.8-STABLE i386) MIME-Version: 1.0 To: Alex Rousskov Cc: freebsd-net@FreeBSD.ORG Subject: Re: TCP throughput vs number of aliases References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Alex Rousskov wrote: > Hi there, > > We use FreeBSD 3.1 to benchmark Web caching proxies. We want to > use a [very] large number of IP addresses for simulated clients and > servers in the benchmark setup. However, we are limited by the number of > machines we can dedicate to the tests. Naturally, we are using interface > aliases to create IP addresses. > > With the number of aliases above 600+ we are monitoring a sudden > and significant drop in TCP throughput (measured using simple netperf > tests). The graph and some details of the setup are at > > http://www.ircache.net/~wessels/Junk/aliases/ > > We would like to use more than 600 aliases; probably 2,000-5,000 aliases > per box without the loss of performance. > > Can anybody recommend a solution? > > Thank you, > > Alex. > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message The code in ip_input.c seems to step through the list of all ip addresses for that interface. It is probably stepping thro' all the alias addresses I wonder if that is the problem. jayanth To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 17:57:10 1999 Delivered-To: freebsd-net@freebsd.org Received: from coconut.itojun.org (coconut.itojun.org [210.160.95.97]) by hub.freebsd.org (Postfix) with ESMTP id F052714FE2 for ; Tue, 26 Oct 1999 17:57:06 -0700 (PDT) (envelope-from itojun@itojun.org) Received: from kiwi.itojun.org (localhost.itojun.org [127.0.0.1]) by coconut.itojun.org (8.9.3+3.2W/3.7W) with ESMTP id JAA15753; Wed, 27 Oct 1999 09:56:57 +0900 (JST) To: jayanth Cc: Alex Rousskov , freebsd-net@FreeBSD.ORG In-reply-to: jayanth's message of Tue, 26 Oct 1999 17:35:24 MST. <381648CC.B57E06BF@yahoo-inc.com> X-Template-Reply-To: itojun@itojun.org X-Template-Return-Receipt-To: itojun@itojun.org X-PGP-Fingerprint: F8 24 B4 2C 8C 98 57 FD 90 5F B4 60 79 54 16 E2 Subject: Re: TCP throughput vs number of aliases From: itojun@iijlab.net Date: Wed, 27 Oct 1999 09:56:57 +0900 Message-ID: <15751.940985817@coconut.itojun.org> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >The code in ip_input.c seems to step through the list of all >ip addresses for that interface. It is probably stepping thro' all the >alias addresses >I wonder if that is the problem. One possible solution for this is implemented in KAME IPv6 stack, for IPv6 only. We may want to port that part to IPv4 if it is common to have more than 500 interface addresses onto a host. If you really are interested you can grab the code from www.kame.net, or http://www2.kame.net/dev/cvsweb.cgi/kame/kame/sys/netinet6/ip6_input.c Basically KAME code avoids linear search over list of interface addresses, and use routing table lookup to check if the packet is destined to ourselves. If your host is assigned 203.178.141.194/27, You would have an entry like below. You can perform routing table lookup against ip->ip_dst, and see if this is routed to loopback interface to know if the packet is toward me. >203.178.141.192/27 link#1 UC 0 0 >203.178.141.194 0:80:ad:71:81:fc UHLW 1 4351966 lo0 KAME team needed to do this for IPv6 because IPv6 has 128bit address and # of comparison instruction would become 4 times more than IPv4 case. The change actually needs some more twists, and is obviously incompatible with past code. We may leave the current linear search code in the source, and use routing table-based lookup as shortcut, like: 1. lookup routing table, if it looks to be ours, goto ours. (ADDED) 2. linear search over list of interface addresses, if it looks to be ours, goto ours. 3. if I'm a router, forward it. 4. drop it. By commenting (1) out, you get 100% backward compatible code/behavior. itojun To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Oct 26 21:12:31 1999 Delivered-To: freebsd-net@freebsd.org Received: from lestat.nas.nasa.gov (lestat.nas.nasa.gov [129.99.33.127]) by hub.freebsd.org (Postfix) with ESMTP id 17A8414CAA for ; Tue, 26 Oct 1999 21:12:29 -0700 (PDT) (envelope-from thorpej@lestat.nas.nasa.gov) Received: from lestat (localhost [127.0.0.1]) by lestat.nas.nasa.gov (8.8.8/8.6.12) with ESMTP id VAA08719; Tue, 26 Oct 1999 21:11:53 -0700 (PDT) Message-Id: <199910270411.VAA08719@lestat.nas.nasa.gov> To: itojun@iijlab.net Cc: jayanth , Alex Rousskov , freebsd-net@FreeBSD.ORG Subject: Re: TCP throughput vs number of aliases Reply-To: Jason Thorpe From: Jason Thorpe Date: Tue, 26 Oct 1999 21:11:52 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Wed, 27 Oct 1999 09:56:57 +0900 itojun@iijlab.net wrote: > > >The code in ip_input.c seems to step through the list of all > >ip addresses for that interface. It is probably stepping thro' all the > >alias addresses > >I wonder if that is the problem. > > One possible solution for this is implemented in KAME IPv6 stack, > for IPv6 only. We may want to port that part to IPv4 if it is common > to have more than 500 interface addresses onto a host. Thor Simon solved this problem in NetBSD quite some time ago by putting in_ifaddrs in a hash table. His application was, in fact, a web server with hundreds of IP addresses. -- Jason R. Thorpe To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 0:18: 0 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id 36CAC14F66 for ; Wed, 27 Oct 1999 00:17:57 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id AAA18992; Wed, 27 Oct 1999 00:17:55 -0700 (PDT) Message-ID: <205001bf204c$4113e750$0201010a@fuzzer.com> From: "Craig Critchley" To: Cc: "Garrett Wollman" , References: <199910270021.RAA23991@implode.root.com> Subject: Re: FTP Net Performance Date: Wed, 27 Oct 1999 00:24:06 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org From: David Greenman > >Looking at netstat -i, I see that I'm getting Oerrs of about 15% of Opkts, > >but no Ierrs. There is an inconsequential number of collisions. Perhaps > >the error rate is to blame for the problem. What sorts of things do these > >errors represent? > > You shouldn't be getting any output errors - that is definately indicating > a problem. Have you tried replacing the cable? What sort of motherboard is > the card plugged into? I've tried three or four different cables, all gave about the same performance, and I still see increasing Oerrs (more like 25%, actually) every time I FTP from the BSD machine. The machine is home-assembled based on an Asus P2L97 with a K6-200 and a Netgear FA310TX PCI netcard. For historical reasons, I've got a Kingston netcard sitting in the machine as well; tomorrow I will try rearranging the ip configuration to use that interface. (I tried this with and without the Kingston driver linked in the kernel, but haven't tried switching to use it.) I should probably pull extraneous cards as well... I should have probably exhausted such things before mailing the experts; as a bsd newbie, I was hoping there was a "oh, yeah, that's almost always X" type answer that I couldn't find in the faq or archives. Sorry if that's a nuisance. Thanks for the help, and if you think of anything else, please let me know. ...Craig To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 8:25:29 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 3FB5914FFE for ; Wed, 27 Oct 1999 08:25:25 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id LAA87666; Wed, 27 Oct 1999 11:25:22 -0400 (EDT) (envelope-from wollman) Date: Wed, 27 Oct 1999 11:25:22 -0400 (EDT) From: Garrett Wollman Message-Id: <199910271525.LAA87666@khavrinen.lcs.mit.edu> To: "Craig Critchley" Cc: "Garrett Wollman" , Subject: Re: FTP Net Performance In-Reply-To: <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> References: <1df701bf1f84$32751020$0201010a@fuzzer.com> <199910261606.MAA83232@khavrinen.lcs.mit.edu> <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Looking at netstat -i, I see that I'm getting Oerrs of about 15% of Opkts, > but no Ierrs. There is an inconsequential number of collisions. Perhaps > the error rate is to blame for the problem. What sorts of things do these > errors represent? TCP performance is going to suck with that level of packet loss. TCP only works well with less than about 1% packet loss; if you're losing 15%, that explains what's going on. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 8:46:25 1999 Delivered-To: freebsd-net@freebsd.org Received: from guard.polynet.lviv.ua (Guard.PolyNet.Lviv.UA [209.58.62.194]) by hub.freebsd.org (Postfix) with SMTP id 8EE1015359 for ; Wed, 27 Oct 1999 08:46:00 -0700 (PDT) (envelope-from pam@postoffice.polynet.lviv.ua) Received: (qmail 2238 invoked from network); 27 Oct 1999 15:45:56 -0000 Received: from unknown (HELO postoffice.polynet.lviv.ua) (unknown) by unknown with SMTP; 27 Oct 1999 15:45:56 -0000 Received: (qmail 95672 invoked by uid 1001); 27 Oct 1999 15:45:55 -0000 Date: 27 Oct 1999 18:45:55 +0300 Message-ID: <19991027184555.A89816@polynet.lviv.ua> Date: Wed, 27 Oct 1999 18:45:55 +0300 From: Adrian Pavlykevych To: freebsd-net@freebsd.org Subject: Problem with 802.1q VLANs under 3.3-Stable Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.2i Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello! Recenty I've tried to configure tagged VLANs on FreeBSD 3.3-Stable. After sucessful compilation of kernel with "pseudo-device vlan 16" option and compiling patched ifconfig with vlan support (from recently published vlan.tar.gz tarball by Bill Paul ), I've configured two VLANs on the xl0 interface and connected it to 3Com SuperStack Switch 3300. I'm able to ping hosts in both VLANs (and get appropriate vlan_start message from kernel), but no UDP or TCP traffic seem to go through. BTW, I've seen similar message in the freebsd-* archives but no real solution/bugfix. Can you advice on the ways to debug this? (I know that tcpdump is not VLAN capable and ends with kernel panic). Answering hosts were connected to VLANs without tagging so each have seen only one VLAN. The only devices doing tag insertion were FreeBSD and SuperStack. Same config with Linux 2.2-something with VLAN support works ok. I've tested with both stock if_vlan.c and one from Pauls tarball. I'll try to provide you with any debug information upon request. Card: 3Com 905B Switch: 3Com SuperStack || Switch 3300 FreeBSD: 3.3-Stable CVSupped on Oct 16 Thanks in advance, -- Adrian Pavlykevych email: System Administrator phone/fax: +380 (322) 742041 State University "Lvivska Polytechnica" To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 15:40:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from pail.ircache.net (pail.scd.ucar.edu [128.117.28.5]) by hub.freebsd.org (Postfix) with ESMTP id 1284E14A11 for ; Wed, 27 Oct 1999 15:40:39 -0700 (PDT) (envelope-from rousskov@ircache.net) Received: from localhost (rousskov@localhost) by pail.ircache.net (8.9.2/8.8.7) with ESMTP id QAA51102 for ; Wed, 27 Oct 1999 16:40:39 -0600 (MDT) (envelope-from rousskov@ircache.net) Date: Wed, 27 Oct 1999 16:40:39 -0600 (MDT) From: Alex Rousskov To: net@freebsd.org Subject: Re: TCP throughput vs number of aliases In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Thanks a lot for everybody who has replied! Based on the your pointers and suggestions, we added a hash to speedup the linear search in ip_input.c. Preliminary results show 94Mbps throughput with 5000 aliases; the same throughput that we were getting with no aliases. I am not sure if our ugly (but very small) hack is of any value to anybody else. The patches suggested on this list are all of better quality, but were either too rigid and/or too big for our purposes. Thanks again, Alex. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 21:33:56 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail.bsb.nutecnet.com.br (mail.bsb.nutecnet.com.br [200.252.253.5]) by hub.freebsd.org (Postfix) with ESMTP id 274CF14C8F for ; Wed, 27 Oct 1999 21:33:39 -0700 (PDT) (envelope-from visi0n@aux-tech.org) Received: from variola.chinatown.org (dl7216-bsb.bsb.nutecnet.com.br [200.252.208.216]) by mail.bsb.nutecnet.com.br (8.8.5/SCA-6.6) with SMTP id CAA18346; Thu, 28 Oct 1999 02:32:58 -0200 (BRV) Date: Thu, 28 Oct 1999 02:35:18 +0000 ( ) From: visi0n X-Sender: visi0n@variola.chinatown.org Reply-To: visi0n To: Scott Hess Cc: Garrett Wollman , freebsd-net@FreeBSD.ORG Subject: Re: RFC 2140. In-Reply-To: <051301bf1f10$82064510$1e80000a@avantgo.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org There's a team for this "project" ?? =============================================================================== visi0n AUX TECHNOLOGIES www.aux-tech.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Oct 27 22:52:14 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id 954AA14E44 for ; Wed, 27 Oct 1999 22:52:12 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id WAA05729; Wed, 27 Oct 1999 22:52:10 -0700 (PDT) Message-ID: <216001bf2109$718c8500$0201010a@fuzzer.com> From: "Craig Critchley" To: "Garrett Wollman" , References: <1df701bf1f84$32751020$0201010a@fuzzer.com><199910261606.MAA83232@khavrinen.lcs.mit.edu><1f6601bf200e$c8e476b0$0201010a@fuzzer.com> <199910271525.LAA87666@khavrinen.lcs.mit.edu> Subject: Re: FTP Net Performance Date: Wed, 27 Oct 1999 22:58:22 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Ok, all the extra stuff is out of the machine, so it's down to motherboard, video card, and netcard. Still getting the same lousy throughput. I took another look at tcpdump, not the graphs, but the data itself. It appears that Windows is sort of behind in its acks and eventually acks twice with the same sequence number. When this happens, everybody holds their breath for a second and a half, then the BSD machine resends the double-acked packet. This is my first tcpdump analysis effort, so if I have misinterpreted something, please let me know. I just ran tcpdump with no switches after setting up BPF and captured the output. If this rings any bells with anyone, I would love to hear your theory. Unfortunately, I only have the one BSD machine, so I can't easily test whether this is restricted to Windows clients. I'm not enough of a TCP expert to know whether the sequence below breaks the rules. Here's a representative sample. I've trimmed down the timestamps and machine names and took off the window and flag information. The windows on both ends were constant. There were no other packets between these, TCP or otherwise. #the server sends some data: 17.55 fuzz.ftp-data > win.1786: . 281781:283241(1460) ack 1 17.55 fuzz.ftp-data > win.1786: . 283241:284701(1460) ack 1 17.55 fuzz.ftp-data > win.1786: . 284701:286161(1460) ack 1 #Windows acks the middle one 17.55 win.1786 > fuzz.ftp-data: . ack 283241 17.55 fuzz.ftp-data > win.1786: . 286161:287621(1460) ack 1 #Windows acks the third 17.55 win.1786 > fuzz.ftp-data: . ack 284701 17.55 fuzz.ftp-data > win.1786: . 287621:289081(1460) ack 1 #Hmm... Windows ack's the same number! 17.55 win.1786 > fuzz.ftp-data: . ack 284701 #Oh no! Big delay, then resend of double acked packet. 19.05 fuzz.ftp-data > win.1786: . 284701:286161(1460) ack 1 ...Craig From: Garrett Wollman > > TCP performance is going to suck with that level of packet loss. TCP > only works well with less than about 1% packet loss; if you're losing > 15%, that explains what's going on. > > -GAWollman > > -- > Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same > wollman@lcs.mit.edu | O Siem / The fires of freedom > Opinions not those of| Dance in the burning flame > MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick > > > 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 From owner-freebsd-net Thu Oct 28 0: 5:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from implode.root.com (root.com [209.102.106.178]) by hub.freebsd.org (Postfix) with ESMTP id 47D8314E65 for ; Thu, 28 Oct 1999 00:05:24 -0700 (PDT) (envelope-from dg@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.8/8.8.5) with ESMTP id AAA27351; Thu, 28 Oct 1999 00:01:51 -0700 (PDT) Message-Id: <199910280701.AAA27351@implode.root.com> To: "Craig Critchley" Cc: "Garrett Wollman" , net@FreeBSD.ORG Subject: Re: FTP Net Performance In-reply-to: Your message of "Wed, 27 Oct 1999 22:58:22 PDT." <216001bf2109$718c8500$0201010a@fuzzer.com> From: David Greenman Reply-To: dg@root.com Date: Thu, 28 Oct 1999 00:01:51 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >Ok, all the extra stuff is out of the machine, so it's down to motherboard, >video card, and netcard. > >Still getting the same lousy throughput. > >I took another look at tcpdump, not the graphs, but the data itself. It >appears that Windows is sort of behind in its acks and eventually acks twice >with the same sequence number. When this happens, everybody holds their >breath for a second and a half, then the BSD machine resends the >double-acked packet. This is my first tcpdump analysis effort, so if I have >misinterpreted something, please let me know. I just ran tcpdump with no >switches after setting up BPF and captured the output. > >If this rings any bells with anyone, I would love to hear your theory. Uh, yeah, that very clearly shows that the Windows box is dropping packets occasionally. Probably insufficient buffering (or just lousy network card) on the Windows side. -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project - http://www.freebsd.org Creator of high-performance Internet servers - http://www.terasolutions.com Pave the road of life with opportunities. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 0:32:48 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id BC05014E65 for ; Thu, 28 Oct 1999 00:32:46 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id AAA11340; Thu, 28 Oct 1999 00:32:44 -0700 (PDT) Message-ID: <217001bf2117$7dfb5150$0201010a@fuzzer.com> From: "Craig Critchley" To: Cc: "Garrett Wollman" , References: <199910280701.AAA27351@implode.root.com> Subject: Re: FTP Net Performance Date: Thu, 28 Oct 1999 00:38:56 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org From: David Greenman > Uh, yeah, that very clearly shows that the Windows box is dropping packets > occasionally. Probably insufficient buffering (or just lousy network card) on > the Windows side. It's the same model netcard as what's in the BSD machine - Netgear FA310TX. File sharing between Windows machines is *significantly* faster. I can imagine that BSD thinks its already sent as much as the receiver can take, but why does it think it needs to resend that particular piece when Windows has ack'd it twice (unless doing so is a bad idea - like I said, I'm not that much of an expert). Thanks, ...Craig To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 0:36:41 1999 Delivered-To: freebsd-net@freebsd.org Received: from jade.chc-chimes.com (jade.chc-chimes.com [216.28.46.6]) by hub.freebsd.org (Postfix) with ESMTP id 33C2F14F7C for ; Thu, 28 Oct 1999 00:36:28 -0700 (PDT) (envelope-from billf@chc-chimes.com) Received: by jade.chc-chimes.com (Postfix, from userid 1001) id E45761C2B; Thu, 28 Oct 1999 02:38:52 -0400 (EDT) Received: from localhost (localhost [127.0.0.1]) by jade.chc-chimes.com (Postfix) with ESMTP id D672C3837; Thu, 28 Oct 1999 02:38:52 -0400 (EDT) Date: Thu, 28 Oct 1999 02:38:52 -0400 (EDT) From: Bill Fumerola To: Craig Critchley Cc: dg@root.com, Garrett Wollman , net@FreeBSD.ORG Subject: Re: FTP Net Performance In-Reply-To: <217001bf2117$7dfb5150$0201010a@fuzzer.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Thu, 28 Oct 1999, Craig Critchley wrote: > It's the same model netcard as what's in the BSD machine - Netgear FA310TX. > File sharing between Windows machines is *significantly* faster. I can > imagine that BSD thinks its already sent as much as the receiver can take, > but why does it think it needs to resend that particular piece when Windows > has ack'd it twice (unless doing so is a bad idea - like I said, I'm not > that much of an expert). ACKing twice is an indictation of dropped packets. It says "i'm going to keep acknowledging the last known good packet until you send me the one after it." .. or something like that. -- - bill fumerola - billf@chc-chimes.com - BF1560 - computer horizons corp - - ph:(800) 252-2421 - bfumerol@computerhorizons.com - billf@FreeBSD.org - To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 1: 1:19 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp.nwlink.com (smtp.nwlink.com [209.20.130.57]) by hub.freebsd.org (Postfix) with ESMTP id 82DC214CFD for ; Thu, 28 Oct 1999 01:01:17 -0700 (PDT) (envelope-from craigc@nwlink.com) Received: from craigc (ip133.gte8.rb1.bel.nwlink.com [209.20.237.133]) by smtp.nwlink.com (8.9.3/8.9.3) with SMTP id BAA12590; Thu, 28 Oct 1999 01:01:13 -0700 (PDT) Message-ID: <218001bf211b$791005b0$0201010a@fuzzer.com> From: "Craig Critchley" To: "Bill Fumerola" Cc: , "Garrett Wollman" , References: Subject: Re: FTP Net Performance Date: Thu, 28 Oct 1999 01:07:26 -0700 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Ah, I've misunderstood the sequence number being ack'd - it's the next byte expected by the reciever, eh? Damn. I've probably been tearing apart the wrong machine. I still don't understand the 1.5 sec delay before BSD resends, though. ...Craig From: Bill Fumerola > On Thu, 28 Oct 1999, Craig Critchley wrote: > > > It's the same model netcard as what's in the BSD machine - Netgear FA310TX. > > File sharing between Windows machines is *significantly* faster. I can > > imagine that BSD thinks its already sent as much as the receiver can take, > > but why does it think it needs to resend that particular piece when Windows > > has ack'd it twice (unless doing so is a bad idea - like I said, I'm not > > that much of an expert). > > ACKing twice is an indictation of dropped packets. It says "i'm going to > keep acknowledging the last known good packet until you send me the one > after it." > > .. or something like that. > > -- > - bill fumerola - billf@chc-chimes.com - BF1560 - computer horizons corp - > - ph:(800) 252-2421 - bfumerol@computerhorizons.com - billf@FreeBSD.org - > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 1:59:22 1999 Delivered-To: freebsd-net@freebsd.org Received: from penguin.wise.edt.ericsson.se (penguin-ext.wise.edt.ericsson.se [194.237.142.110]) by hub.freebsd.org (Postfix) with ESMTP id F14E414E19 for ; Thu, 28 Oct 1999 01:59:11 -0700 (PDT) (envelope-from jacopo.pecci@erv.ericsson.se) Received: from puh.eritel.se (puh.erv.ericsson.se [194.22.140.20]) by penguin.wise.edt.ericsson.se (8.9.3/8.9.3/WIREfire-1.5) with ESMTP id KAA00642 for ; Thu, 28 Oct 1999 10:59:09 +0200 (MET DST) Received: from esegsnt003.erv.ericsson.se (esegsnt003 [194.22.140.6]) by puh.eritel.se (8.8.8+Sun/8.8.8) with ESMTP id KAA24478 for ; Thu, 28 Oct 1999 10:58:38 +0200 (MET DST) Received: by esegsnt003.erv.ericsson.se with Internet Mail Service (5.5.2650.21) id ; Thu, 28 Oct 1999 10:58:38 +0200 Message-ID: <2DADC88F8F19D31193340090277A3661017920F7@esegsnt003.erv.ericsson.se> From: Jacopo Pecci To: "'FREEBSD'" Subject: TCP threshold Date: Thu, 28 Oct 1999 10:58:27 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I am experiment some problem with the variable "ssthresh" (the threshold that marks the change from Slow start to Congestion avoidance).in TCP, FreeBSD. I have just discovered that every time I start a new socket the value of this variable depends on the final value in the previous socket. It means that if socket 1 has some drop of packets, the next socket I will establish, will have a much lover value of the threshold. WHY? each socket is completely independent from the previous, why it should inherit such information?. I have repeated 7 times one experiment that encountered 1 drop each time and at the end the starting value of this variable was 26K, pretty small compare with the very first value 1G! I would like to do several measurements and compare the results but I can not do that if I always change my scenario. Is there any way to avoid that? (the only way I know is to reboot the machine every time) Could I change TCP code or that might be dangerous? Thanks, Jacopo. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 2: 0:58 1999 Delivered-To: freebsd-net@freebsd.org Received: from implode.root.com (root.com [209.102.106.178]) by hub.freebsd.org (Postfix) with ESMTP id 0346014F8E for ; Thu, 28 Oct 1999 02:00:51 -0700 (PDT) (envelope-from dg@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.8/8.8.5) with ESMTP id BAA27530; Thu, 28 Oct 1999 01:57:50 -0700 (PDT) Message-Id: <199910280857.BAA27530@implode.root.com> To: "Craig Critchley" Cc: net@FreeBSD.ORG Subject: Re: FTP Net Performance In-reply-to: Your message of "Thu, 28 Oct 1999 01:07:26 PDT." <218001bf211b$791005b0$0201010a@fuzzer.com> From: David Greenman Reply-To: dg@root.com Date: Thu, 28 Oct 1999 01:57:50 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >Ah, I've misunderstood the sequence number being ack'd - it's the next byte >expected by the reciever, eh? Damn. I've probably been tearing apart the >wrong machine. > >I still don't understand the 1.5 sec delay before BSD resends, though. That's the minimum retransmit timeout. -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project - http://www.freebsd.org Creator of high-performance Internet servers - http://www.terasolutions.com Pave the road of life with opportunities. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 7:42: 4 1999 Delivered-To: freebsd-net@freebsd.org Received: from cdc.net (server1.cdc.net [207.244.0.12]) by hub.freebsd.org (Postfix) with SMTP id CC5B914D46 for ; Thu, 28 Oct 1999 07:41:58 -0700 (PDT) (envelope-from mwade@cdc.net) Received: (qmail 25371 invoked by uid 100); 28 Oct 1999 14:41:54 -0000 Date: Thu, 28 Oct 1999 10:41:54 -0400 (EDT) From: Mike Wade X-Sender: mwade@server1 To: net@FreeBSD.ORG Subject: RIPv2/OSPF over tun0 using GateD? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Is it possible to use GateD with RIPv2/OSPF over the tunnel interfaces? ie: tun0 and friends? If so, how? --- Mike Wade (mwade@cdc.net) Director of Systems Administration CDC Internet, Inc. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 8: 2:29 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id EA44C14D46 for ; Thu, 28 Oct 1999 08:02:26 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id LAA91565; Thu, 28 Oct 1999 11:01:08 -0400 (EDT) (envelope-from wollman) Date: Thu, 28 Oct 1999 11:01:08 -0400 (EDT) From: Garrett Wollman Message-Id: <199910281501.LAA91565@khavrinen.lcs.mit.edu> To: Jacopo Pecci Cc: "'FREEBSD'" Subject: TCP threshold In-Reply-To: <2DADC88F8F19D31193340090277A3661017920F7@esegsnt003.erv.ericsson.se> References: <2DADC88F8F19D31193340090277A3661017920F7@esegsnt003.erv.ericsson.se> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > I have just discovered that every time I start a new socket the value of > this variable depends on the final value in the previous socket. It means > that if socket 1 has some drop of packets, the next socket I will establish, > will have a much lover value of the threshold. WHY? each socket is > completely independent from the previous, why it should inherit such > information?. Because you are connecting to the same destination; the conditions in the network are unlikely to vary significantly over the (short) timescale during which the information is cached. You can disable the caching by using the `route' command to set a lock on the ``ssthresh'' route metric for the route(s) in question. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 8: 7:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 925A614CF1 for ; Thu, 28 Oct 1999 08:07:24 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id LAA91582; Thu, 28 Oct 1999 11:07:22 -0400 (EDT) (envelope-from wollman) Date: Thu, 28 Oct 1999 11:07:22 -0400 (EDT) From: Garrett Wollman Message-Id: <199910281507.LAA91582@khavrinen.lcs.mit.edu> To: "Craig Critchley" Cc: Subject: Re: FTP Net Performance In-Reply-To: <218001bf211b$791005b0$0201010a@fuzzer.com> References: <218001bf211b$791005b0$0201010a@fuzzer.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Ah, I've misunderstood the sequence number being ack'd - it's the next byte > expected by the reciever, eh? Damn. I've probably been tearing apart the > wrong machine. No, it's the sequence number of the last octet (or control bit) successfully received *and desegmented*. The SYN and FIN bits occupy one unit in sequence space. TCP's acknowledgements are cumulative, so ACK(1234) means that all data up to and including sequence number 1234 have been received successfully. If sequence 1222 is lost, then regardless of how many later packets are successfully received, TCP will send ACK(1221) until a segment containing 1222 is received. In any event, duplicate acknowledgements are a sign of two possible problems: 1) Packets are getting re-ordered (not likely in your configuration). 2) Packets are getting dropped (which, as you already stated, was happening about 25% of the time). TCP will always perform poorly when large numbers of packets are being dropped, because it must wait for a retransmission timeout (minimum value 0.5 seconds in -current and 1.5 +/- 0.5 seconds in -stable) and then slow-start. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 8: 9:13 1999 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 2603414CF1 for ; Thu, 28 Oct 1999 08:09:09 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id LAA91590; Thu, 28 Oct 1999 11:09:07 -0400 (EDT) (envelope-from wollman) Date: Thu, 28 Oct 1999 11:09:07 -0400 (EDT) From: Garrett Wollman Message-Id: <199910281509.LAA91590@khavrinen.lcs.mit.edu> To: "Craig Critchley" Cc: Subject: Re: FTP Net Performance In-Reply-To: <216001bf2109$718c8500$0201010a@fuzzer.com> References: <1df701bf1f84$32751020$0201010a@fuzzer.com> <199910261606.MAA83232@khavrinen.lcs.mit.edu> <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> <199910271525.LAA87666@khavrinen.lcs.mit.edu> <216001bf2109$718c8500$0201010a@fuzzer.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > I took another look at tcpdump, not the graphs, but the data itself. It would have been much easier to identify the problem by looking at a plot. I've already explained what's going on -- an output error is indicative of a dropped packet. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 8:32: 1 1999 Delivered-To: freebsd-net@freebsd.org Received: from nimitz.ca.sandia.gov (nimitz.ca.sandia.gov [146.246.243.56]) by hub.freebsd.org (Postfix) with ESMTP id C43ED15103 for ; Thu, 28 Oct 1999 08:31:53 -0700 (PDT) (envelope-from bmah@nimitz.ca.sandia.gov) Received: (from bmah@localhost) by nimitz.ca.sandia.gov (8.9.3/8.9.3) id IAA97032; Thu, 28 Oct 1999 08:31:46 -0700 (PDT) Message-Id: <199910281531.IAA97032@nimitz.ca.sandia.gov> X-Mailer: exmh version 2.1.0 09/18/1999 To: Garrett Wollman Cc: "Craig Critchley" , net@FreeBSD.ORG Subject: Re: FTP Net Performance In-Reply-To: Your message of "Thu, 28 Oct 1999 11:07:22 EDT." <199910281507.LAA91582@khavrinen.lcs.mit.edu> From: bmah@CA.Sandia.GOV (Bruce A. Mah) Reply-To: bmah@CA.Sandia.GOV X-Face: g~c`.{#4q0"(V*b#g[i~rXgm*w;:nMfz%_RZLma)UgGN&=j`5vXoU^@n5v4:OO)c["!w)nD/!!~e4Sj7LiT'6*wZ83454H""lb{CC%T37O!!'S$S&D}sem7I[A 2V%N&+ X-Url: http://www.ca.sandia.gov/~bmah/ Mime-Version: 1.0 Content-Type: multipart/signed; boundary="==_Exmh_-804607468P"; micalg=pgp-sha1; protocol="application/pgp-signature" Content-Transfer-Encoding: 7bit Date: Thu, 28 Oct 1999 08:31:46 -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org --==_Exmh_-804607468P Content-Type: text/plain; charset=us-ascii If memory serves me right, Garrett Wollman wrote: > < s > aid: > > > Ah, I've misunderstood the sequence number being ack'd - it's the next byte > > expected by the reciever, eh? Damn. I've probably been tearing apart the > > wrong machine. > > No, it's the sequence number of the last octet (or control bit) > successfully received *and desegmented*. The SYN and FIN bits occupy > one unit in sequence space. TCP's acknowledgements are cumulative, so > ACK(1234) means that all data up to and including sequence number 1234 > have been received successfully. If sequence 1222 is lost, then > regardless of how many later packets are successfully received, TCP > will send ACK(1221) until a segment containing 1222 is received. Possibly contributing to the confusion is the way that tcpdump prints out TCP sequence numbers. If you have a line that looks like... 08:22:19.252316 foo.ssh > bar.935: P 1:37(36) ack 20 win 8760 (DF) [tos 0x10] ...the segment displayed really contained octets with sequence numbers 1 through 36 inclusive, despite the fact that the line reads "1:37" (36 objects numbered starting from 1 will end at 36, not 37). Note that the next segment from this example conversation was... 08:22:20.472454 foo.ssh > bar.935: P 37:73(36) ack 40 win 8760 (DF) [tos 0x10] ...showing that byte 37 really went in the second segment, not the first. If a person didn't know this fact, it'd be easy to interpret the acks wrongly. Bruce. --==_Exmh_-804607468P Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 5.0i for non-commercial use MessageID: PxW59vcMMOAxFckYIAHWJKhNVEf9nMri iQA/AwUBOBhsYtjKMXFboFLDEQLB+wCfT89VgdmYJKcoubXQp9TK1MWX2AoAoOeY UrTtKo16RwOBIwqgY2DcjztA =AwR2 -----END PGP SIGNATURE----- --==_Exmh_-804607468P-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 15: 7:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from awfulhak.org (dynamic-86.max1-du-ws.dialnetwork.pavilion.co.uk [212.74.8.86]) by hub.freebsd.org (Postfix) with ESMTP id 3D8BD14C49 for ; Thu, 28 Oct 1999 15:07:20 -0700 (PDT) (envelope-from brian@Awfulhak.org) Received: from hak.lan.Awfulhak.org (root@hak.lan.Awfulhak.org [172.16.0.12]) by awfulhak.org (8.9.3/8.9.3) with ESMTP id WAA01468; Thu, 28 Oct 1999 22:59:27 +0100 (BST) (envelope-from brian@lan.awfulhak.org) Received: from hak.lan.Awfulhak.org (brian@localhost.lan.Awfulhak.org [127.0.0.1]) by hak.lan.Awfulhak.org (8.9.3/8.9.3) with ESMTP id TAA00484; Thu, 28 Oct 1999 19:10:07 +0100 (BST) (envelope-from brian@hak.lan.Awfulhak.org) Message-Id: <199910281810.TAA00484@hak.lan.Awfulhak.org> X-Mailer: exmh version 2.1.0 09/18/1999 To: Julian Elischer Cc: Denis DeLaRoca , net@FreeBSD.ORG, brian@hak.lan.Awfulhak.org Subject: Re: PPPOE report. In-Reply-To: Message from Julian Elischer of "Tue, 26 Oct 1999 01:44:59 PDT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 28 Oct 1999 19:10:07 +0100 From: Brian Somers Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > yes there is news.. > > The PPPOE node negotiated a session on the test setup and I'm just > cleaning up some stuff. > > Brian is doing some work on the ppp code to support it and when I'm > finished doing this I'll send him some support code. I think I'm just about there now.... there'll be a commit to -current soon. [.....] -- Brian Don't _EVER_ lose your sense of humour ! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Oct 28 15:22: 5 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpo.whistle.com (alpo.whistle.com [207.76.204.38]) by hub.freebsd.org (Postfix) with ESMTP id 47DC514CB3 for ; Thu, 28 Oct 1999 15:22:02 -0700 (PDT) (envelope-from julian@whistle.com) Received: from current1.whiste.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with ESMTP id PAA49959; Thu, 28 Oct 1999 15:21:52 -0700 (PDT) Date: Thu, 28 Oct 1999 15:21:52 -0700 (PDT) From: Julian Elischer To: Brian Somers Cc: Denis DeLaRoca , net@FreeBSD.ORG, brian@hak.lan.Awfulhak.org Subject: Re: PPPOE report. In-Reply-To: <199910281810.TAA00484@hak.lan.Awfulhak.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org brian I'm about to try test this.. (just applying patches) gather you are on-line.. I'll throw a window on freefall and if you want we can 'talk" julian On Thu, 28 Oct 1999, Brian Somers wrote: > > yes there is news.. > > > > The PPPOE node negotiated a session on the test setup and I'm just > > cleaning up some stuff. > > > > Brian is doing some work on the ppp code to support it and when I'm > > finished doing this I'll send him some support code. > > I think I'm just about there now.... there'll be a commit to -current > soon. > > [.....] > -- > Brian > > Don't _EVER_ lose your sense of humour ! > > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Oct 29 6:32:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from dent.axion.bt.co.uk (dent.axion.bt.co.uk [132.146.16.161]) by hub.freebsd.org (Postfix) with ESMTP id 5AE5C155D4 for ; Fri, 29 Oct 1999 06:32:18 -0700 (PDT) (envelope-from antonio.herrera-alcantara@bt.com) Received: from cbtlipnt01.btlabs.bt.co.uk by dent (local) with ESMTP; Fri, 29 Oct 1999 14:29:59 +0100 Received: by cbtlipnt01.btlabs.bt.co.uk with Internet Mail Service (5.5.2448.0) id ; Fri, 29 Oct 1999 14:29:10 +0100 Message-ID: <97E01B170FC1D211B8EB0000F8FE9E077116C6@mbtlipnt03.btlabs.bt.co.uk> From: antonio.herrera-alcantara@bt.com To: freebsd-net@freebsd.org Subject: 2 questions Date: Fri, 29 Oct 1999 14:29:07 +0100 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello all, I've got a couple of questions, perhaps some of you can give the answers. The first problem is with a FreeBSD-3.3 machine, it has xl0, fxp0, fxp1, fxp2, fxp3 as network interfaces. It has been configured as a router and I have included the following statements in /etc/rc.conf: gateway_enable="YES" network_interfaces="xl0 fxp0 fxp1 fxp2 fxp3 lo0" ifconfig_xl0="inet 132.146.130.108 netmask 255.255.255.0" ifconfig_fxp0="inet 10.10.30.252 netmask 255.255.255.0" ifconfig_fxp1="inet 10.10.35.252 netmask 255.255.255.0" ifconfig_fxp2="inet 10.10.31.252 netmask 255.255.255.0" ifconfig_fxp3="inet 10.10.23.252 netmask 255.255.255.0" defaultrouter="132.146.130.254" After rebooting the routing table doen't have entries for fxp0, ..., fxp3, only for xl0 and lo0. I've got tow other machines with a very similar configuration and this problem doesn't happen. Furthermoe, if I manully delete the IP address from fxp(0,1,2,3) with ifconfig and I reassing it the routing table gets updated. What's wrong?. Just after rebooting, netstat -rn gives: Routing tables Internet: Destination Gateway Flags Refs Use Netif Expire default 132.146.130.254 UGSc 1 88 xl0 127.0.0.1 127.0.0.1 UH 0 2 lo0 132.146.130/24 link#3 UC 0 0 xl0 While it should be something like: Routing tables Internet: Destination Gateway Flags Refs Use Netif Expire default 132.146.130.254 UGSc 1 88 xl0 10.10.23/24 link#5 UC 0 0 fxp3 10.10.30/24 link#1 UC 0 0 fxp0 10.10.31/24 link#4 UC 0 0 fxp2 10.10.35/24 link#2 UC 0 0 fxp1 127.0.0.1 127.0.0.1 UH 0 2 lo0 132.146.130/24 link#3 UC 0 0 xl0 Second question: if I want to add two static routes at start-up in a host, for instance: route add -net 10.10.31.0 10.10.11.254 route add -net 10.10.12.0 10.10.11.254 What is the systanx to achieve that using the static routes list within /etc/rc.conf: static_routes="??????????" Thanks in advance for your help, Antonio Herrera. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Oct 29 10:37:38 1999 Delivered-To: freebsd-net@freebsd.org Received: from nippon.highcaliber.com (nippon.highcaliber.com [206.217.210.2]) by hub.freebsd.org (Postfix) with ESMTP id 54F8E152DA for ; Fri, 29 Oct 1999 10:37:36 -0700 (PDT) (envelope-from Andre@HighCaliber.com) Received: from uranus ([206.217.210.24]) by nippon.highcaliber.com (post.office MTA v1.9.3 ID# 0-16273) with SMTP id AAA199 for ; Fri, 29 Oct 1999 13:39:52 -0400 Message-ID: <004e01bf2235$16e7e310$18d2d9ce@uranus.highcaliber.com> From: Andre@HighCaliber.com (Andre Chang) To: Subject: Date: Fri, 29 Oct 1999 13:43:20 -0400 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3612.1700 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3612.1700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org subscribe To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Oct 29 11: 4: 6 1999 Delivered-To: freebsd-net@freebsd.org Received: from nippon.highcaliber.com (nippon.highcaliber.com [206.217.210.2]) by hub.freebsd.org (Postfix) with ESMTP id 0628A1502E for ; Fri, 29 Oct 1999 11:04:03 -0700 (PDT) (envelope-from Andre@HighCaliber.com) Received: from uranus ([206.217.210.24]) by nippon.highcaliber.com (post.office MTA v1.9.3 ID# 0-16273) with SMTP id AAA203 for ; Fri, 29 Oct 1999 14:06:18 -0400 Message-ID: <008201bf2238$c83f7c60$18d2d9ce@uranus.highcaliber.com> From: Andre@HighCaliber.com (Andre Chang) To: Subject: Third Ethernet card "fxp2" Date: Fri, 29 Oct 1999 14:09:46 -0400 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.3612.1700 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3612.1700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello, I'd like to know if there are any known issues when enabling a 3rd ethernet interface on a system config'd as follows 1 PR440FX 2 Pentium Pro Processors 128MB RAM SCSI HDD and CDROM 3 Intel EtherExpress Pro 10/100B adapters (fxp0 is onboard) 3.3-STABLE FreeBSD 3.3-STABLE #3: Sat Oct 23 15:57:22 EST 1999 SMP Kernel IPFW enabled NATD enabled Specifically, when I enable the fxp2 interface via rc.conf by adding these 2 lines: ifconfig_fxp2="inet 192.168.20.1 netmask 255.255.255.0" network_interfaces="fxp0 fxp1 fxp2 lo0" The machine hangs when attempting to ifconfig fxp2 I got around the problem by physically removing the card and the system booted normally with the expected error that fxp2 does not exist. Any help would be greatly appreciated. I just subscribed to this group so please cc me also. Thanks. -- Andre. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Oct 29 18:27:47 1999 Delivered-To: freebsd-net@freebsd.org Received: from granite.sentex.net (granite.sentex.ca [199.212.134.1]) by hub.freebsd.org (Postfix) with ESMTP id F3F2D14C3A for ; Fri, 29 Oct 1999 18:27:29 -0700 (PDT) (envelope-from mike@sentex.net) Received: from ospf-mdt.sentex.net (ospf-mdt.sentex.net [205.211.164.81]) by granite.sentex.net (8.8.8/8.6.9) with SMTP id VAA27251; Fri, 29 Oct 1999 21:27:26 -0400 (EDT) From: mike@sentex.net (Mike Tancsa) To: Andre@HighCaliber.com (Andre Chang) Cc: freebsd-net@freebsd.org Subject: Re: Third Ethernet card "fxp2" Date: Sat, 30 Oct 1999 01:27:25 GMT Message-ID: <381a2d12.436676377@mail.sentex.net> References: In-Reply-To: X-Mailer: Forte Agent .99e/32.227 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 29 Oct 1999 14:06:46 -0400, in sentex.lists.freebsd.net you wrote: >Hello, > >I'd like to know if there are any known issues when enabling a 3rd ethernet >interface on a system config'd as follows > >1 PR440FX >2 Pentium Pro Processors >128MB RAM >SCSI HDD and CDROM >3 Intel EtherExpress Pro 10/100B adapters (fxp0 is onboard) > >3.3-STABLE FreeBSD 3.3-STABLE #3: Sat Oct 23 15:57:22 EST 1999 >SMP Kernel >IPFW enabled >NATD enabled There shouldnt be.. On one of my machines, grep fxp /etc/rc.conf network_interfaces="fxp0 fxp1 fxp2 fxp3 lo0" A couple of things to check perhaps, try and make sure the cards all have their individual IRQs, and that if there are no IDE drives in the box, disable the IDE controllers so that they dont take up IRQs and DMA resources. What sort of SCSI controller do you have ? ---Mike Mike Tancsa (mdtancsa@sentex.net) Sentex Communications Corp, Waterloo, Ontario, Canada "Given enough time, 100 monkeys on 100 routers could setup a national IP network." (KDW2) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Oct 29 23:43:22 1999 Delivered-To: freebsd-net@freebsd.org Received: from dfssl.exchange.microsoft.com (dfssl.exchange.microsoft.com [131.107.88.59]) by hub.freebsd.org (Postfix) with ESMTP id 587A814F66; Fri, 29 Oct 1999 23:43:16 -0700 (PDT) (envelope-from minwei@Exchange.Microsoft.com) Received: by dfssl with Internet Mail Service (5.5.2650.21) id ; Fri, 29 Oct 1999 23:43:11 -0700 Message-ID: <01D6C7224936D211BA450000F805D5380F8D04B4@TOTO> From: "Min Wei (Exchange)" To: freebsd-net@FreeBSD.ORG, "'hackers@freebsd.org'" Subject: FreeBSD reboots every 45 minutes Date: Fri, 29 Oct 1999 23:41:31 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01BF22A1.CA8731C0" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF22A1.CA8731C0 Content-Type: text/plain Hi. I am confused at the following strange FreeBSD reboot case. I have a Dell PowerEdge 2300 dual proc machine with Intel 100/10 netcard. I put 3.2 SMP kernel with options MAXUSERS=256 and NMBCLUSTERS=32768. Otherwise kernel is almost the same as the stock FreeBSD kernel. When I started some live Web tests with my custom Apache module, I found that the FreeBSD panics every 45 minutes. First, I thought it was some FIN_WAIT_2 problem. So I applied julian(@whistle.com)'s patch. But it still reboots after 45 minutes of testing. Then I thought it might be running out of mbufs due to load. However, the number of allocated mbufs is no where near 32768 just seconds before the reboot. One additional piece of data is that in my simluated environment, I was able to pound the machine with FreeBSD 3.2 SMP for more than 12 hours straight with 600 running Apache child processes. I haven't tried to dump kernel core to see the call stack since the machine is actually far away at a hosting site. I wonder if anyone is aware of SMP kernel issues with Dell PowerEdge machines. I will try single processor mode when I could arrange more live time. However I am kind of suspect somewhere in net driver or TCP/IP stack is screwed up. Any help on the possible reboot cause is greatly appreciated. Otherwise I might have to try BSDI which is something I would rather not to do. Thanks, --min ------_=_NextPart_001_01BF22A1.CA8731C0 Content-Type: text/html FreeBSD reboots every 45 minutes

Hi. I am confused at the following strange FreeBSD reboot case.
I have a Dell PowerEdge 2300 dual proc machine with Intel 100/10 netcard.
I put 3.2 SMP kernel with options MAXUSERS=256 and NMBCLUSTERS=32768.
Otherwise kernel is almost the same as the stock FreeBSD kernel.

When I started some live Web tests with my custom Apache module,
I found that the FreeBSD panics every 45 minutes. First, I thought
it was some FIN_WAIT_2 problem. So I applied julian(@whistle.com)'s
patch. But it still reboots after 45 minutes of testing. Then
I thought it might be running out of mbufs due to load. However, the
number of allocated mbufs is no where near 32768 just seconds before the reboot.

One additional piece of data is that in my simluated environment, I was able
to pound the machine with FreeBSD 3.2 SMP for more than 12 hours straight
with 600 running Apache child processes.

I haven't tried to dump kernel core to see the call stack since the machine
is actually far away at a hosting site.

I wonder if anyone is aware of SMP kernel issues with Dell PowerEdge machines.
I will try single processor mode when I could arrange more live time.
However I am kind of suspect somewhere in net driver or TCP/IP stack
is screwed up.

Any help on the possible reboot cause is greatly appreciated. Otherwise I might
have to try BSDI which is something I would rather not to do.

        Thanks,
        --min

------_=_NextPart_001_01BF22A1.CA8731C0-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message