From owner-freebsd-net Sun Mar 21 11:39:54 1999 Delivered-To: freebsd-net@freebsd.org Received: from friley-185-206.res.iastate.edu (friley-185-206.res.iastate.edu [129.186.185.206]) by hub.freebsd.org (Postfix) with ESMTP id 291FB14E93 for ; Sun, 21 Mar 1999 11:39:51 -0800 (PST) (envelope-from cc@137.org) Received: from friley-185-205.res.iastate.edu (friley-185-205.res.iastate.edu [129.186.185.205]) by friley-185-206.res.iastate.edu (Postfix) with ESMTP id DDBF267; Sun, 21 Mar 1999 13:39:30 -0600 (CST) Received: from friley-185-205.res.iastate.edu (localhost [127.0.0.1]) by friley-185-205.res.iastate.edu (Postfix) with ESMTP id 19005C3; Sun, 21 Mar 1999 13:39:30 -0600 (CST) X-Mailer: exmh version 2.0.2 2/24/98 To: Julian Elischer Cc: Chris Csanady , freebsd-net@FreeBSD.ORG Subject: Re: Integrating the NetBSD PFIL hooks.. In-reply-to: Your message of "Fri, 19 Mar 1999 17:36:29 PST." <36F2FB9D.2C67412E@whistle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 21 Mar 1999 13:39:30 -0600 From: Chris Csanady Message-Id: <19990321193930.19005C3@friley-185-205.res.iastate.edu> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >Chris Csanady wrote: >> >> What would it take for us to intergrate NetBSD's PFIL hooks? It is >> hard to do much work in the current network stack with so much of >> the mess that currently exists. At the very least, ip_input.c and >> ip_output.c would be much cleaner with this mechanism. >> >> I'm just wondering what needs to be done, and if it is possible. >> Ipfilter would already support this, but how about ipfw, dummynet, >> divert and such? Would the authors of the respective code be >> willing to help out with the necessary changes? >> >> Chris Csanady >> >> To Unsubscribe: send mail to majordomo@FreeBSD.org >> with "unsubscribe freebsd-net" in the body of the message > >Certainly >though I haven't looked.. >It certainly looks like it could use some cleaning.. It's suffering >from 'evolutionary changes'. > >We at whistle have to take a lot of the blame. >We implemented 'divert' sockets after a suggestion from >one of the CSRG guys. (forget his name.. the Kieth that was >not a Bostic) > >The divert functionality adds a lot of possibilities but it has its >tentacles all over the place. The 'fwd' option of ipfw has a few >tentacles reaching as far as tcp_input. Hmm, I didn't realize that divert was so far reaching. The NetBSD PFIL stuff basically only provides for input and output hooks at a single point as far as I can see. (in ip_input and ip_output) It seems like it would be simple to extend the interface to do both fragments and reassembled packets at the IP layer though. What is the minimum in terms of filtering points that must exist? Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 16:41:41 1999 Delivered-To: freebsd-net@freebsd.org Received: from ATLANTICO.mail.telepac.pt (mail1.telepac.pt [194.65.3.53]) by hub.freebsd.org (Postfix) with ESMTP id 1A1111513D for ; Sun, 21 Mar 1999 16:41:35 -0800 (PST) (envelope-from jpedras@mail.telepac.pt) Received: from manecao.tafkap.priv ([194.65.206.156]) by ATLANTICO.mail.telepac.pt (Intermail v3.1 117 241) with ESMTP id <19990322004116.CIHA16009@manecao.tafkap.priv> for ; Mon, 22 Mar 1999 00:41:16 +0000 Content-Length: 2890 Message-ID: X-Mailer: XFMail 1.3 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 Date: Mon, 22 Mar 1999 00:40:00 -0000 (GMT) From: Joao Pedras To: freebsd-net@freebsd.org Subject: autodisconnect failing Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello everyone! I connect to the internet using one of my FreeBSD boxes using isdn with i4b as a gateway. It always as been configured to disconnect after some idle time. Well it as always been working... until a couple days ago I noticed that it wasn't disconnecting. I have also noticed this problem with several other people using other operating systems in this area and with this isp. The thing is, if I stay with everything idle I am able to notice that there is incoming traffic. The time interval between these packets is 1 or 2 seconds. Please notice that this ONLY happens when the communications to the internet are idle. These packets do NOT show up when normal traffic is taking place. Why am I receiving this packets ? I tried to know what are the contents of those packets and here is the result : Frame (64 on wire, 64 captured) Frame arrived on Mar 22, 1999 00:25:24.1818 Total frame length: 64 bytes Capture frame length: 64 bytes Point-to-Point Protocol Address: ff Control: 03 Protocol: IP (0x0021) Internet Protocol Version: 4 Header length: 20 bytes Type of service: 0x00 (None) 000. .... = routine precedence ...0 .... = normal delay .... 0... = normal throughput .... .0.. = normal reliability .... ..0. = normal cost Total length: 60 Identification: 0x0000 Flags: 0x0 .0.. .... = may fragment ..0. .... = last fragment Fragment offset: 0 Time to live: 2 Protocol: Unknown (58) Header checksum: 0xd274 Source address: coi-rd1.telepac.net (194.65.66.234) Destination address: IGRP-ROUTERS.MCAST.NET (224.0.0.10) Frame (64 on wire, 64 captured) Frame arrived on Mar 22, 1999 00:25:24.1818 Total frame length: 64 bytes Capture frame length: 64 bytes Point-to-Point Protocol Address: ff Control: 03 Protocol: IP (0x0021) Internet Protocol Version: 4 Header length: 20 bytes Type of service: 0x00 (None) 000. .... = routine precedence ...0 .... = normal delay .... 0... = normal throughput .... .0.. = normal reliability .... ..0. = normal cost Total length: 60 Identification: 0x0000 Flags: 0x0 .0.. .... = may fragment ..0. .... = last fragment Fragment offset: 0 Time to live: 2 Protocol: Unknown (58) Header checksum: 0xd274 Source address: coi-rd1.telepac.net (194.65.66.234) Destination address: IGRP-ROUTERS.MCAST.NET (224.0.0.10) Thank you for your help! Joao Pedras --------------------------------------------------- Sent using XFMail on 22-Mar-99 at 00:21:31 This message was sent by XFMail proudly powered by FreeBSD -> http://www.freebsd.org <- "The Power to Serve" --------------------------------------------------- In Devon, Connecticut, it is unlawful to walk backwards after sunset. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 20:28:11 1999 Delivered-To: freebsd-net@freebsd.org Received: from gjp.erols.com (alex-va-n008c079.moon.jic.com [206.156.18.89]) by hub.freebsd.org (Postfix) with ESMTP id 191B1150D3 for ; Sun, 21 Mar 1999 20:28:09 -0800 (PST) (envelope-from gjp@gjp.erols.com) Received: from gjp.erols.com (localhost.erols.com [127.0.0.1]) by gjp.erols.com (8.9.1/8.8.7) with ESMTP id XAA51319; Sun, 21 Mar 1999 23:27:43 -0500 (EST) (envelope-from gjp@gjp.erols.com) X-Mailer: exmh version 2.0.1 12/23/97 To: Julian Elischer Cc: Christopher Sedore , "'freebsd-net@freebsd.org'" From: "Gary Palmer" Subject: Re: clustering/load balancing In-reply-to: Your message of "Thu, 18 Mar 1999 18:46:32 PST." <36F1BA88.2F1CF0FB@whistle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 21 Mar 1999 23:27:43 -0500 Message-ID: <51315.922076863@gjp.erols.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Julian Elischer wrote in message ID <36F1BA88.2F1CF0FB@whistle.com>: > +-------[Machine B] > | > [internet]-----[ Machine A ]-----+-------[Machine C] > | > +-------[Machine D] ^^^^^^^^^ Single Point Of Failure Bad Julian No Biscuit :) Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 20:37:26 1999 Delivered-To: freebsd-net@freebsd.org Received: from obie.softweyr.com (unknown [204.68.178.33]) by hub.freebsd.org (Postfix) with ESMTP id 7D126150D3; Sun, 21 Mar 1999 20:37:09 -0800 (PST) (envelope-from wes@softweyr.com) Received: from softweyr.com (wes@zaphod.softweyr.com [204.68.178.35]) by obie.softweyr.com (8.8.8/8.8.8) with ESMTP id VAA29202; Sun, 21 Mar 1999 21:36:51 -0700 (MST) (envelope-from wes@softweyr.com) Message-ID: <36F5C8E3.F0BDD8C5@softweyr.com> Date: Sun, 21 Mar 1999 21:36:51 -0700 From: Wes Peters Organization: Softweyr llc X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Gary Palmer Cc: Julian Elischer , Christopher Sedore , "'freebsd-net@freebsd.org'" Subject: Re: clustering/load balancing References: <51315.922076863@gjp.erols.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Gary Palmer wrote: > > Julian Elischer wrote in message ID > <36F1BA88.2F1CF0FB@whistle.com>: > > +-------[Machine B] > > | > > [internet]-----[ Machine A ]-----+-------[Machine C] > > | > > +-------[Machine D] > > ^^^^^^^^^ > > Single Point Of Failure "Failure is not an option." ;^) -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC http://www.softweyr.com/~softweyr wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 22: 8:20 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 9BB2914D5C; Sun, 21 Mar 1999 22:08:16 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id EAA04887; Mon, 22 Mar 1999 04:52:49 +0100 From: Luigi Rizzo Message-Id: <199903220352.EAA04887@labinfo.iet.unipi.it> Subject: Re: clustering/load balancing To: gpalmer@FreeBSD.ORG (Gary Palmer) Date: Mon, 22 Mar 1999 04:52:48 +0100 (MET) Cc: julian@whistle.com, cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG In-Reply-To: <51315.922076863@gjp.erols.com> from "Gary Palmer" at Mar 21, 99 11:27:24 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 612 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > Julian Elischer wrote in message ID > <36F1BA88.2F1CF0FB@whistle.com>: > > +-------[Machine B] > > | > > [internet]-----[ Machine A ]-----+-------[Machine C] > > | > > +-------[Machine D] > > ^^^^^^^^^ > > Single Point Of Failure I'd rather say > [internet]-----[ Machine A ]-----+-------[Machine C] > | > +-------[Machine D] ^^^^^^^^^ Single Point Of Failure cheers luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 23: 6:49 1999 Delivered-To: freebsd-net@freebsd.org Received: from agata.clio.it (unknown [195.60.136.3]) by hub.freebsd.org (Postfix) with SMTP id 55B221506D for ; Sun, 21 Mar 1999 23:06:39 -0800 (PST) (envelope-from delphi@agata.clio.it) Received: (qmail 4995 invoked by uid 7770); 22 Mar 1999 07:14:36 -0000 Received: from unknown (HELO agata.clio.it) (195.60.136.16) by 195.60.136.3 with SMTP; 22 Mar 1999 07:14:36 -0000 Message-ID: <36F5ED1E.9F1484B4@agata.clio.it> Date: Mon, 22 Mar 1999 08:11:26 +0100 From: Jilani Khaldi X-Mailer: Mozilla 4.5 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-net@FreeBSD.ORG Subject: (no subject) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org auth dcd0b446 subscribe freebsd-net delphi@agata.clio.it To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sun Mar 21 23:50:58 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.seidata.com (ns1.seidata.com [208.10.211.2]) by hub.freebsd.org (Postfix) with ESMTP id B4F2E14BE9; Sun, 21 Mar 1999 23:50:56 -0800 (PST) (envelope-from mike@seidata.com) Received: from localhost (mike@localhost) by ns1.seidata.com (8.8.8/8.8.5) with ESMTP id CAA00269; Mon, 22 Mar 1999 02:50:38 -0500 (EST) Date: Mon, 22 Mar 1999 02:50:37 -0500 (EST) From: To: Gary Palmer Cc: Julian Elischer , Christopher Sedore , "'freebsd-net@freebsd.org'" Subject: Re: clustering/load balancing In-Reply-To: <51315.922076863@gjp.erols.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 Sun, 21 Mar 1999, Gary Palmer wrote: > Single Point Of Failure Well... I've looked at a lot of clustering technology, reverse proxying, etc. and find it ammusing how the SPF is typically just 'shifted around' rather than eliminated. Heh... SPF will drive you to drinking... Orange Juice, at least. Later, -Mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 0: 7:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from main.piter.net (main.piter.net [195.201.22.10]) by hub.freebsd.org (Postfix) with ESMTP id C466F14CEC; Mon, 22 Mar 1999 00:07:41 -0800 (PST) (envelope-from cyril@main.piter.net) Received: (from cyril@localhost) by main.piter.net (8.8.7/8.8.7/sply) id LAA06347; Mon, 22 Mar 1999 11:07:20 +0300 (MSK) (envelope-from cyril) Date: Mon, 22 Mar 1999 11:07:20 +0300 (MSK) From: "Cyril A. Vechera" Message-Id: <199903220807.LAA06347@main.piter.net> To: gpalmer@FreeBSD.ORG, julian@whistle.com Subject: Re: clustering/load balancing Cc: cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > From owner-freebsd-net@FreeBSD.ORG Mon Mar 22 07:28:07 1999 > To: Julian Elischer > Cc: Christopher Sedore , > "'freebsd-net@freebsd.org'" > From: "Gary Palmer" > Subject: Re: clustering/load balancing > Date: Sun, 21 Mar 1999 23:27:43 -0500 > > Julian Elischer wrote in message ID > <36F1BA88.2F1CF0FB@whistle.com>: > > +-------[Machine B] > > | > > [internet]-----[ Machine A ]-----+-------[Machine C] > > | > > +-------[Machine D] > > ^^^^^^^^^ > > Single Point Of Failure in the original scheme 'single point of failure' is still present. +-------[Machine B] | [internet]-----[ any router ]----+-------[Machine C] | +-------[Machine D] ^^^^^^^^^ or maybe you can see other way to connect 'internet' to Machine [B-C]? what is the differense between 'any router' failures and 'balance dispatcher' failures? Sincerely your, Cyril A. Vechera email:cyril@piter.net --------- http://sply.piter.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 1:21: 3 1999 Delivered-To: freebsd-net@freebsd.org Received: from rose.niw.com.au (app3022-2.gw.connect.com.au [203.63.119.4]) by hub.freebsd.org (Postfix) with ESMTP id 5BA3B15218 for ; Mon, 22 Mar 1999 01:20:47 -0800 (PST) (envelope-from ian@apdata.com.au) Received: from apdata.com.au (localhost [127.0.0.1]) by rose.niw.com.au (Postfix) with ESMTP id 7D5C6A3203 for ; Mon, 22 Mar 1999 19:50:21 +1030 (CST) Message-ID: <36F60B55.1EEFB167@apdata.com.au> Date: Mon, 22 Mar 1999 19:50:21 +1030 From: Ian West Organization: Applied Data Control X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 4.0-CURRENT i386) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-net@freebsd.org Subject: Re: clustering/load balancing References: <199903220807.LAA06347@main.piter.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org "Cyril A. Vechera" wrote: > > > From owner-freebsd-net@FreeBSD.ORG Mon Mar 22 07:28:07 1999 > > To: Julian Elischer > > Cc: Christopher Sedore , > > "'freebsd-net@freebsd.org'" > > From: "Gary Palmer" > > Subject: Re: clustering/load balancing > > Date: Sun, 21 Mar 1999 23:27:43 -0500 > > > > Julian Elischer wrote in message ID > > <36F1BA88.2F1CF0FB@whistle.com>: > > > +-------[Machine B] > > > | > > > [internet]-----[ Machine A ]-----+-------[Machine C] > > > | > > > +-------[Machine D] > > > > ^^^^^^^^^ > > > > Single Point Of Failure > > in the original scheme 'single point of failure' is still present. > > +-------[Machine B] > | > [internet]-----[ any router ]----+-------[Machine C] > | > +-------[Machine D] > ^^^^^^^^^ > > or maybe you can see other way to connect 'internet' to Machine [B-C]? > > what is the differense between 'any router' failures and 'balance > dispatcher' failures? > > Sincerely your, > Cyril A. Vechera > > email:cyril@piter.net --------- http://sply.piter.net > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-net" in the body of the message I have a very similar problem, I am not so much interested in load balancing as fault tolerance. I have been looking at implementing vrrp, which being standard (almost...) would allow a bsd machine to provide redundancy for a router, or allow two bsd machines to provide redundancy for each other. It does not provide load balancing that I can see, but possibly with judicious use of forwarding this could be achieved. The url for the latest draft I am aware of is below http://www.ietf.org/internet-drafts/draft-ietf-vrrp-spec-v2-01.txt I have not spent a lot of tie on it yet, but as far as I can tell, it should not require much other than writing a daemon which would implement the polling/advertising, load the multicast bits into relevant interface(s), and put the actual 'virtual' router address onto a loopback interface. (The loopback seems like the easiest way to make the ip address respond to traffic, but not to arp requests, without changing kernel stuff.) These are my thought to date, and hopefully I will be able to start writing something in a couple of weeks. It is more than possible that I have overlooked something enormous which will be a real show stopper, but this is a function I need, and it may suit quite a few applications. Any comments ? Regards, Ian West To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 2:27:48 1999 Delivered-To: freebsd-net@freebsd.org Received: from rt2.synx.com (tech.boostworks.com [194.167.81.239]) by hub.freebsd.org (Postfix) with ESMTP id 3E0DE14BD3 for ; Mon, 22 Mar 1999 02:26:30 -0800 (PST) (envelope-from root@synx.com) Received: from synx.com (rn.synx.com [192.1.1.241]) by rt2.synx.com (8.9.1/8.9.1) with ESMTP id LAA25950; Mon, 22 Mar 1999 11:32:59 +0100 (CET) Message-Id: <199903221032.LAA25950@rt2.synx.com> Date: Mon, 22 Mar 1999 11:24:41 +0100 (CET) From: Remy Nonnenmacher Reply-To: remy@synx.com Subject: Re: Integrating the NetBSD PFIL hooks.. To: cc@137.org Cc: julian@whistle.com, freebsd-net@FreeBSD.ORG In-Reply-To: <19990321193930.19005C3@friley-185-205.res.iastate.edu> MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 21 Mar, Chris Csanady wrote: > >>Chris Csanady wrote: >>> >>> What would it take for us to intergrate NetBSD's PFIL hooks? It is >>> hard to do much work in the current network stack with so much of >>> the mess that currently exists. At the very least, ip_input.c and >>> ip_output.c would be much cleaner with this mechanism. >>> >>> I'm just wondering what needs to be done, and if it is possible. >>> Ipfilter would already support this, but how about ipfw, dummynet, >>> divert and such? Would the authors of the respective code be >>> willing to help out with the necessary changes? >>> >>> Chris Csanady >>> >>> To Unsubscribe: send mail to majordomo@FreeBSD.org >>> with "unsubscribe freebsd-net" in the body of the message >> >>Certainly >>though I haven't looked.. >>It certainly looks like it could use some cleaning.. It's suffering >>from 'evolutionary changes'. >> >>We at whistle have to take a lot of the blame. >>We implemented 'divert' sockets after a suggestion from >>one of the CSRG guys. (forget his name.. the Kieth that was >>not a Bostic) >> >>The divert functionality adds a lot of possibilities but it has its >>tentacles all over the place. The 'fwd' option of ipfw has a few >>tentacles reaching as far as tcp_input. > > Hmm, I didn't realize that divert was so far reaching. The NetBSD > PFIL stuff basically only provides for input and output hooks at > a single point as far as I can see. (in ip_input and ip_output) > It seems like it would be simple to extend the interface to do > both fragments and reassembled packets at the IP layer though. > > What is the minimum in terms of filtering points that must exist? > > Chris All these 'tentacles' comes from the fact that every point of processing of a packet needs to know more about interest that this packet can have to an upper-layer. Even bridging in the sentence : "bridge all packets but forward this connection to this socket" must be able to understand quiet everything, from Mac to pcb in order to route the packet. What about generalizing the whistle's Netgraph to the full stack and building a generic traffic classifier ? (I bothered Luigi about it a few time... Sorry Luigi ;). RN. IaM To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 3:15:21 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 2278F14CC3; Mon, 22 Mar 1999 03:15:20 -0800 (PST) (envelope-from julian@whistle.com) Received: from current1.whistle.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with SMTP id DAA29144; Mon, 22 Mar 1999 03:10:13 -0800 (PST) Date: Mon, 22 Mar 1999 03:10:13 -0800 (PST) From: Julian Elischer To: Gary Palmer Cc: Christopher Sedore , "'freebsd-net@freebsd.org'" Subject: Re: clustering/load balancing In-Reply-To: <51315.922076863@gjp.erols.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 so? I was just giving an example of load sharing using exisiting code. I probably wouldn't have A actually doing work, that way it'd be a hell of a lot more reliable, and hey, a PC is cheap..have another on standby. On Sun, 21 Mar 1999, Gary Palmer wrote: > Julian Elischer wrote in message ID > <36F1BA88.2F1CF0FB@whistle.com>: > > +-------[Machine B] > > | > > [internet]-----[ Machine A ]-----+-------[Machine C] > > | > > +-------[Machine D] > > ^^^^^^^^^ > > Single Point Of Failure > > Bad Julian > > No Biscuit > > :) > > Gary > -- > Gary Palmer FreeBSD Core Team Member > FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info > > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 7:29:12 1999 Delivered-To: freebsd-net@freebsd.org Received: from maxwell.syr.edu (maxwell.syr.edu [128.230.129.5]) by hub.freebsd.org (Postfix) with ESMTP id EC27F15168 for ; Mon, 22 Mar 1999 07:27:41 -0800 (PST) (envelope-from cmsedore@maxwell.syr.edu) Received: from exchange.maxwell.syr.edu (exchange.maxwell.syr.edu [128.230.129.241]) by maxwell.syr.edu (8.9.1a/8.9.1) with ESMTP id KAA02888; Mon, 22 Mar 1999 10:14:44 GMT Received: by exchange.maxwell.syr.edu with Internet Mail Service (5.5.1960.3) id ; Mon, 22 Mar 1999 10:27:10 -0500 Message-ID: <262C3DA9BE0CD211971700A0C9B413A1CBDD@exchange.maxwell.syr.edu> From: Christopher Sedore To: "'Cyril A. Vechera'" Cc: freebsd-net@FreeBSD.ORG Subject: RE: clustering/load balancing Date: Mon, 22 Mar 1999 10:27:03 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.1960.3) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > -----Original Message----- > From: Cyril A. Vechera [mailto:cyril@main.piter.net] > Subject: Re: clustering/load balancing > > > > From: "Gary Palmer" > > Subject: Re: clustering/load balancing > > > > Julian Elischer wrote in message ID > > <36F1BA88.2F1CF0FB@whistle.com>: > > > +-------[Machine B] > > > | > > > [internet]-----[ Machine A ]-----+-------[Machine C] > > > | > > > +-------[Machine D] > > > > ^^^^^^^^^ > > > > Single Point Of Failure > > in the original scheme 'single point of failure' is still present. > > +-------[Machine B] > | > [internet]-----[ any router ]----+-------[Machine C] > | > +-------[Machine D] > ^^^^^^^^^ > > or maybe you can see other way to connect 'internet' to Machine [B-C]? > > what is the differense between 'any router' failures and 'balance > dispatcher' failures? The difference is that you can have multiple links to the Internet or whatever other network you're trying to serve with your cluster. There are ways to achieve this with multiple dispatch machines, but its not as appealing a solution. -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 8: 9:55 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail.dynamixweb.com (host01.dynamixweb.com [209.47.109.194]) by hub.freebsd.org (Postfix) with ESMTP id 05D001511A for ; Mon, 22 Mar 1999 08:09:49 -0800 (PST) (envelope-from svetzal@icom.ca) Received: from blazer (cr609409-a.pr1.on.wave.home.com [24.112.98.34]) by mail.dynamixweb.com with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2232.9) id H2VCMAJG; Sun, 21 Mar 1999 09:14:03 -0500 From: "Steven Vetzal" To: Subject: RE: natd and ipmasq Date: Sun, 21 Mar 1999 09:11:45 -0500 Message-ID: <000001be73a4$c15422e0$7ffea8c0@blazer.pr1.on.wave.home.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook 8.5, Build 4.71.2173.0 In-Reply-To: X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Importance: Normal Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > what's the difference between NATD dan ipmasquerading on LInux ? > i thought that was the same.. is that true ? I won't go much into natd vs. IP Masquerading, I see a couple people beat me to the punch, but... I wondered often about the functional differences between the two for quite some time until I ran across someone running IP Masquerading on Linux (haven't had the time to swap my gateway at home from FreeBSD to Linux to check it out). After looking into it, I discovered a few things. Mainly, the core IP masquerading function in Linux seems to require several patches to allow things like ping and traceroute to work through the gateway, whereas on FreeBSD these things have been rolled into the core natd process. In fact, recently on this list Ari was kind enough to point me towards the natd development site, where I successfully compiled natd 2.0beta to allow GRE (a tunnelling protocol) to also work through the FreeBSD gateway I'm using. Again, to do this on Linux required either a couple patches and a proxy program, or other wierdness. Basically, as I've always found, FreeBSD has proven to be a far more straightforward and stable platform for many purposes, firewalling and such. I've been running it at home for about a year now on my cablemodem, and it's been awesome. Best of all, I didn't have to spend $40K on a Cisco PIX 8^) But I did know how to do PPTP through Linux IPMasquerading long before I found out how to do PPTP through FreeBSD natd 8^) Steve To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 8:41:34 1999 Delivered-To: freebsd-net@freebsd.org Received: from gjp.erols.com (alex-va-n008c079.moon.jic.com [206.156.18.89]) by hub.freebsd.org (Postfix) with ESMTP id 5BE9F1528D for ; Mon, 22 Mar 1999 08:41:05 -0800 (PST) (envelope-from gjp@gjp.erols.com) Received: from gjp.erols.com (localhost.erols.com [127.0.0.1]) by gjp.erols.com (8.9.1/8.8.7) with ESMTP id LAA60607; Mon, 22 Mar 1999 11:40:33 -0500 (EST) (envelope-from gjp@gjp.erols.com) X-Mailer: exmh version 2.0.1 12/23/97 To: mike@seidata.com Cc: Julian Elischer , Christopher Sedore , "'freebsd-net@freebsd.org'" From: "Gary Palmer" Subject: Re: clustering/load balancing In-reply-to: Your message of "Mon, 22 Mar 1999 02:50:37 EST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 22 Mar 1999 11:40:33 -0500 Message-ID: <60603.922120833@gjp.erols.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org mike@seidata.com wrote in message ID : > Well... I've looked at a lot of clustering technology, reverse > proxying, etc. and find it ammusing how the SPF is typically just > 'shifted around' rather than eliminated. > > Heh... SPF will drive you to drinking... Orange Juice, at least. > > Later, Most load balancing solutions have some sort of hot-failover between redundant switches. Alteons GigE host adapters even have failover between cards (unfortunately done in software). (anyone know of a FastEther NIC with failover?) Combine that with HSRP'd routers and diverse fiber paths and you're coming pretty damned close to not having a SPoF Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 8:46:53 1999 Delivered-To: freebsd-net@freebsd.org Received: from gjp.erols.com (alex-va-n008c079.moon.jic.com [206.156.18.89]) by hub.freebsd.org (Postfix) with ESMTP id 2B9B715256 for ; Mon, 22 Mar 1999 08:46:51 -0800 (PST) (envelope-from gjp@gjp.erols.com) Received: from gjp.erols.com (localhost.erols.com [127.0.0.1]) by gjp.erols.com (8.9.1/8.8.7) with ESMTP id LAA60699; Mon, 22 Mar 1999 11:45:59 -0500 (EST) (envelope-from gjp@gjp.erols.com) X-Mailer: exmh version 2.0.1 12/23/97 To: "Cyril A. Vechera" Cc: julian@whistle.com, cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG From: "Gary Palmer" Subject: Re: clustering/load balancing In-reply-to: Your message of "Mon, 22 Mar 1999 11:07:20 +0300." <199903220807.LAA06347@main.piter.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 22 Mar 1999 11:45:59 -0500 Message-ID: <60695.922121159@gjp.erols.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org "Cyril A. Vechera" wrote in message ID <199903220807.LAA06347@main.piter.net>: > > in the original scheme 'single point of failure' is still present. > > +-------[Machine B] > | > [internet]-----[ any router ]----+-------[Machine C] > | > +-------[Machine D] > ^^^^^^^^^ > > or maybe you can see other way to connect 'internet' to Machine [B-C]? There are many ways of doing it... Just very few ways that remove the SPoF all the way to the host. HSRP'd routers and per-packet load balancing have reduced a lot of the router based failures that could happen, but that still doesn't remove the SPoF in the switch/hub between the routers and the hosts. As I said in a previous message, Alteon have failover between switches, although it probably requires their GigE cards to go to the host and have that work. > what is the differense between 'any router' failures and 'balance > dispatcher' failures? I added a smiley to the end of the message for a reason. There are solutions to most of the other SPoFs that are out there today. The biggest one still left is actually not the one I highlighted, but rather: +-------[Machine B] | [internet]-----[ any router ]----+-------[Machine C] | +-------[Machine D] ^^^^^^^^^^^^ Anyone know of any way to have redudancy all the way to the host? (i.e. 2 or more NICs) Its going to need some daemon on the host watching the NIC for a heartbeat or something, then sending out an ARP invalidation packet for the (now failed) NIC and then another ARP for the (now working) NIC. Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 9:19:14 1999 Delivered-To: freebsd-net@freebsd.org Received: from gjp.erols.com (alex-va-n008c079.moon.jic.com [206.156.18.89]) by hub.freebsd.org (Postfix) with ESMTP id 9D26714D12 for ; Mon, 22 Mar 1999 09:19:12 -0800 (PST) (envelope-from gjp@gjp.erols.com) Received: from gjp.erols.com (localhost.erols.com [127.0.0.1]) by gjp.erols.com (8.9.1/8.8.7) with ESMTP id MAA61120; Mon, 22 Mar 1999 12:18:49 -0500 (EST) (envelope-from gjp@gjp.erols.com) X-Mailer: exmh version 2.0.1 12/23/97 To: Julian Elischer Cc: Christopher Sedore , "'freebsd-net@freebsd.org'" From: "Gary Palmer" Subject: Re: clustering/load balancing In-reply-to: Your message of "Mon, 22 Mar 1999 03:10:13 PST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 22 Mar 1999 12:18:49 -0500 Message-ID: <61116.922123129@gjp.erols.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Julian Elischer wrote in message ID : > so? > I was just giving an example of load sharing using exisiting > code. > I probably wouldn't have A actually doing work, > that way it'd be a hell of a lot more reliable, > and hey, a PC is cheap..have another on standby. I wasn't faulting you Julian... there was a smiley firmly attached. If nothing else, even if you remove that SPoF, you leave others (mainly between the `load balancer' and the actual servers, unless you've figured out a way to have redundant NIC's in your servers) Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 9:34:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from Thanatos.Shenton.Org (Thanatos.Shenton.Org [209.31.147.195]) by hub.freebsd.org (Postfix) with ESMTP id 761A114BD5; Mon, 22 Mar 1999 09:34:24 -0800 (PST) (envelope-from chris@shenton.org) Received: (from chris@localhost) by Thanatos.Shenton.Org (8.9.2/8.9.2) id MAA25174; Mon, 22 Mar 1999 12:40:16 -0500 (EST) (envelope-from chris) To: Julian Elischer Cc: Gary Palmer , Christopher Sedore , "'freebsd-net@freebsd.org'" Subject: Re: clustering/load balancing References: From: Chris Shenton Date: 22 Mar 1999 12:40:13 -0500 In-Reply-To: Julian Elischer's message of "Mon, 22 Mar 1999 03:10:13 -0800 (PST)" Message-ID: <87lngpe83m.fsf@Thanatos.Shenton.Org> Lines: 28 X-Mailer: Gnus v5.6.45/Emacs 20.3 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Julian Elischer writes: > I was just giving an example of load sharing using exisiting code. > I probably wouldn't have A actually doing work, that way it'd be a > hell of a lot more reliable, and hey, a PC is cheap..have another on > standby. I've missed the origin o the thread so feel free to ignore me if I'm missing the point. I don't like manual fail-over and unused hardware waiting for the "real" server to die. At a number of sites, I've used boxes from RND Networks (www.rndnetworks.com) to do server load sharing/balancing. If you use them singly, it looks like the single point o' failure diagram posted here. But you can use them in pairs which test each other for upness and if one goes belly-up the other assumes its IP and MAC so no other machines notice. These are pricey, too pricey IMHO, but so is the competition; Acuitive.com has an interesting review of the technology and products if you're interested. The Ericsson folks have released open source software to do something very similar using a load-balancing DNS server, IP migration, intelligent HTTP gateway, and there's talk of content replication. They say it's in Beta for Linux :-( but have plans to release a Solaris version; would be nice to see it run on FreeBSD. I haven't played with it yet but it looks way cool. It's at www.eddieware.org. If anyone's used it, I'd be interested in a report. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 9:44:38 1999 Delivered-To: freebsd-net@freebsd.org Received: from maxwell.syr.edu (maxwell.syr.edu [128.230.129.5]) by hub.freebsd.org (Postfix) with ESMTP id 285CF14C88; Mon, 22 Mar 1999 09:44:36 -0800 (PST) (envelope-from cmsedore@maxwell.syr.edu) Received: from exchange.maxwell.syr.edu (exchange.maxwell.syr.edu [128.230.129.241]) by maxwell.syr.edu (8.9.1a/8.9.1) with ESMTP id MAA06846; Mon, 22 Mar 1999 12:31:50 GMT Received: by exchange.maxwell.syr.edu with Internet Mail Service (5.5.1960.3) id ; Mon, 22 Mar 1999 12:44:17 -0500 Message-ID: <262C3DA9BE0CD211971700A0C9B413A1CBDE@exchange.maxwell.syr.edu> From: Christopher Sedore To: "'Gary Palmer'" , "Cyril A. Vechera" Cc: julian@whistle.com, Christopher Sedore , freebsd-net@FreeBSD.ORG Subject: RE: clustering/load balancing Date: Mon, 22 Mar 1999 12:44:15 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.1960.3) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > -----Original Message----- > From: Gary Palmer [mailto:gpalmer@FreeBSD.ORG] > Sent: Monday, March 22, 1999 11:46 AM > Subject: Re: clustering/load balancing > > I added a smiley to the end of the message for a reason. > There are solutions > to most of the other SPoFs that are out there today. The > biggest one still > left is actually not the one I highlighted, but rather: > > > +-------[Machine B] > | > [internet]-----[ any router ]----+-------[Machine C] > | > +-------[Machine D] > ^^^^^^^^^^^^ > > Anyone know of any way to have redudancy all the way to the > host? (i.e. 2 or > more NICs) Its going to need some daemon on the host watching > the NIC for a > heartbeat or something, then sending out an ARP invalidation > packet for the > (now failed) NIC and then another ARP for the (now working) NIC. DAS FDDI solves this nicely, right? No hub, and if your redundant switches/routers do FDDI, you've reduced the problem back to a node failure. -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 9:47:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from poboxer.pobox.com (unknown [208.149.16.10]) by hub.freebsd.org (Postfix) with ESMTP id 5F718152DD; Mon, 22 Mar 1999 09:47:40 -0800 (PST) (envelope-from alk@poboxer.pobox.com) Received: (from alk@localhost) by poboxer.pobox.com (8.9.3/8.9.1) id KAA02084; Mon, 22 Mar 1999 10:46:22 -0600 (CST) (envelope-from alk) From: Anthony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 22 Mar 1999 10:46:22 -0600 (CST) X-Face: \h9Jg:Cuivl4S*UP-)gO.6O=T]]@ncM*tn4zG);)lk#4|lqEx=*talx?.Gk,dMQU2)ptPC17cpBzm(l'M|H8BUF1&]dDCxZ.c~Wy6-j,^V1E(NtX$FpkkdnJixsJHE95JlhO 5\M3jh'YiO7KPCn0~W`Ro44_TB@&JuuqRqgPL'0/{):7rU-%.*@/>q?1&Ed Reply-To: alk@pobox.com To: gpalmer@FreeBSD.ORG Cc: julian@whistle.com, cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG Subject: Re: clustering/load balancing References: <61116.922123129@gjp.erols.com> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <14070.29383.300007.474845@avalon.east> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Quoth Gary Palmer on Mon, 22 March: : even if you remove that SPoF, you leave others (mainly between the `load : balancer' and the actual servers, unless you've figured out a way to have : redundant NIC's in your servers) You don't need physically redundant nic in order to have multiple load balancers: The servers don't care whether packets are routed to them through nic 1 or nic 2. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 11:14: 8 1999 Delivered-To: freebsd-net@freebsd.org Received: from gjp.erols.com (alex-va-n008c079.moon.jic.com [206.156.18.89]) by hub.freebsd.org (Postfix) with ESMTP id D605F15074 for ; Mon, 22 Mar 1999 11:14:06 -0800 (PST) (envelope-from gjp@gjp.erols.com) Received: from gjp.erols.com (localhost.erols.com [127.0.0.1]) by gjp.erols.com (8.9.1/8.8.7) with ESMTP id OAA62646; Mon, 22 Mar 1999 14:13:35 -0500 (EST) (envelope-from gjp@gjp.erols.com) To: alk@pobox.com Cc: julian@whistle.com, cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG From: "Gary Palmer" Subject: Re: clustering/load balancing In-reply-to: Your message of "Mon, 22 Mar 1999 10:46:22 CST." <14070.29383.300007.474845@avalon.east> Date: Mon, 22 Mar 1999 14:13:35 -0500 Message-ID: <62642.922130015@gjp.erols.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Anthony Kimball wrote in message ID <14070.29383.300007.474845@avalon.east>: > Quoth Gary Palmer on Mon, 22 March: > : even if you remove that SPoF, you leave others (mainly between the `load > : balancer' and the actual servers, unless you've figured out a way to have > : redundant NIC's in your servers) > > You don't need physically redundant nic in order to have multiple > load balancers: The servers don't care whether packets are routed to > them through nic 1 or nic 2. But how do you route back out? Running gated is one answer I guess... Not one I'd readily choose. Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 11:19:56 1999 Delivered-To: freebsd-net@freebsd.org Received: from poboxer.pobox.com (unknown [208.149.16.10]) by hub.freebsd.org (Postfix) with ESMTP id AE1A214D68; Mon, 22 Mar 1999 11:19:52 -0800 (PST) (envelope-from alk@poboxer.pobox.com) Received: (from alk@localhost) by poboxer.pobox.com (8.9.3/8.9.1) id NAA02536; Mon, 22 Mar 1999 13:19:42 -0600 (CST) (envelope-from alk) From: Anthony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 22 Mar 1999 13:19:42 -0600 (CST) X-Face: \h9Jg:Cuivl4S*UP-)gO.6O=T]]@ncM*tn4zG);)lk#4|lqEx=*talx?.Gk,dMQU2)ptPC17cpBzm(l'M|H8BUF1&]dDCxZ.c~Wy6-j,^V1E(NtX$FpkkdnJixsJHE95JlhO 5\M3jh'YiO7KPCn0~W`Ro44_TB@&JuuqRqgPL'0/{):7rU-%.*@/>q?1&Ed Reply-To: alk@pobox.com To: gpalmer@FreeBSD.ORG Cc: freebsd-net@FreeBSD.ORG Subject: Re: clustering/load balancing References: <14070.29383.300007.474845@avalon.east> <62642.922130015@gjp.erols.com> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <14070.38645.340702.523783@avalon.east> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Quoth Gary Palmer on Mon, 22 March: : > : > You don't need physically redundant nic in order to have multiple : > load balancers: The servers don't care whether packets are routed to : > them through nic 1 or nic 2. : : But how do you route back out? Running gated is one answer I guess... : Not one I'd readily choose. I'm less squeamish. But even so, it's a non-issue for connection-oriented protocols. If they just serve web or nfs over tcp, no worries. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 14:46:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from orbit.flnet.com (orbit.flnet.com [205.240.232.32]) by hub.freebsd.org (Postfix) with ESMTP id 7E2F71559B; Mon, 22 Mar 1999 14:46:24 -0800 (PST) (envelope-from henrich@orbit.flnet.com) Received: (from henrich@localhost) by orbit.flnet.com (8.8.5/8.8.4) id RAA17593; Mon, 22 Mar 1999 17:46:01 -0500 (EST) Date: Mon, 22 Mar 1999 14:46:01 -0800 From: Charles Henrich To: freebsd-hackers@freebsd.org, freebsd-net@freebsd.org Subject: NAT/SKIP/MTU Message-ID: <19990322144600.A17340@orbit.flnet.com> Mail-Followup-To: freebsd-hackers@freebsd.org, freebsd-net@freebsd.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.4i X-Operating-System: FreeBSD 2.2-BETA_A X-PGP-Fingerprint: 1024/F7 FD C7 3A F5 6A 23 BF 76 C4 B8 C9 6E 41 A4 4F Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I've run into ap roblem where Im attempting to do both NAT and SKIP on the same machine... However whenever the MTU of the internal (net 10) interface is less than 1500, packets are either dropped or never reassembled properly causing communication with a variety of internet hosts to be broken. SKIP alters the MTU to 1336 (I'm assuming to make space in the packet for the encryption overhead)... When it does this though, everything goes to hell. Has anyone else out there seen this problem and come up with a solution? Is this a FreeBSD networking issue, or is it a problem with NAT, or even worse, is this a problem with other hosts on the network not being able to cope with a different MTU? (I initially noticed this problem with travelocity, and expedia's web sites..) Thanks for any info! -Crh Charles Henrich Manex Visual Effects henrich@flnet.com http://orbit.flnet.com/~henrich To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Mon Mar 22 23:52:59 1999 Delivered-To: freebsd-net@freebsd.org Received: from main.piter.net (main.piter.net [195.201.22.10]) by hub.freebsd.org (Postfix) with ESMTP id DD21A1563D; Mon, 22 Mar 1999 23:52:55 -0800 (PST) (envelope-from cyril@main.piter.net) Received: (from cyril@localhost) by main.piter.net (8.8.7/8.8.7/sply) id KAA21587; Tue, 23 Mar 1999 10:52:23 +0300 (MSK) (envelope-from cyril) Date: Tue, 23 Mar 1999 10:52:23 +0300 (MSK) From: "Cyril A. Vechera" Message-Id: <199903230752.KAA21587@main.piter.net> To: gpalmer@FreeBSD.ORG, julian@whistle.com Subject: Re: clustering/load balancing Cc: cmsedore@maxwell.syr.edu, freebsd-net@FreeBSD.ORG Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > From owner-freebsd-net@FreeBSD.ORG Mon Mar 22 20:19:19 1999 > To: Julian Elischer > Cc: Christopher Sedore , > "'freebsd-net@freebsd.org'" > From: "Gary Palmer" > Subject: Re: clustering/load balancing > Date: Mon, 22 Mar 1999 12:18:49 -0500 > > Julian Elischer wrote in message ID > : > > so? > > I was just giving an example of load sharing using exisiting > > code. > > I probably wouldn't have A actually doing work, > > that way it'd be a hell of a lot more reliable, > > and hey, a PC is cheap..have another on standby. > > I wasn't faulting you Julian... there was a smiley firmly attached. If nothing > else, even if you remove that SPoF, you leave others (mainly between the `load > balancer' and the actual servers, unless you've figured out a way to have > redundant NIC's in your servers) heh, ok. let's think that we do not need any 'failure safe' technologies, only load balancing. in this way does 'forwarders scheme' is better than 'ARP patched'? or there are other underwater stones? Sincerely your, Cyril A. Vechera email:cyril@piter.net --------- http://sply.piter.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 7:34:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from maxwell.syr.edu (maxwell.syr.edu [128.230.129.5]) by hub.freebsd.org (Postfix) with ESMTP id DC4FA15489 for ; Tue, 23 Mar 1999 07:34:21 -0800 (PST) (envelope-from cmsedore@maxwell.syr.edu) Received: from exchange.maxwell.syr.edu (exchange.maxwell.syr.edu [128.230.129.241]) by maxwell.syr.edu (8.9.1a/8.9.1) with ESMTP id KAA24403 for ; Tue, 23 Mar 1999 10:21:29 GMT Received: by exchange.maxwell.syr.edu with Internet Mail Service (5.5.1960.3) id ; Tue, 23 Mar 1999 10:34:02 -0500 Message-ID: <262C3DA9BE0CD211971700A0C9B413A1CBE1@exchange.maxwell.syr.edu> From: Christopher Sedore To: "'freebsd-net@freebsd.org'" Subject: clustering stuff Date: Tue, 23 Mar 1999 10:34:01 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.1960.3) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I tar'ed up the patches, cluster daemon code, and a readme, and put them up for review and tryout. See http://tfeed.maxwell.syr.edu -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 10:15:43 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail.promo.de (mail.Promo.DE [194.45.188.65]) by hub.freebsd.org (Postfix) with ESMTP id F3C3614DF8 for ; Tue, 23 Mar 1999 10:15:03 -0800 (PST) (envelope-from stefan.bethke@hanse.de) Received: from d225.promo.de (d225.Promo.DE [194.45.188.225]) by mail.promo.de (8.8.8/8.8.8) with ESMTP id TAA16504 for ; Tue, 23 Mar 1999 19:14:37 +0100 (CET) Date: Tue, 23 Mar 1999 19:14:37 +0100 From: Stefan Bethke To: freebsd-net@freebsd.org Subject: natd fails for -interface Message-ID: <153968.3131205277@d225.promo.de> Originator-Info: login-id=stefan; server=mail X-Mailer: Mulberry (MacOS) [1.4.2, s/n U-301178] MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Finally, I found the solution to a nagging problem: For some interfaces, I wasn't able to run natd -interface if0. It started OK, but as the first packet arrived, it said natd: Unknown interface name isp5. (Undefined error: 0) The problem is in SetAliasAddressFromIfName(), where for the SIOCGIFCONF, = a buffer for only 32 entries is allocated: struct ifreq buf[32]; While space for 32 addresses will suffice for a home setup with only two = or three interfaces and a limited number of addresses, I have three machines with multiple ISDN connections which have up to 42 (sic!) interface addresses assigned. I would suggest to up this to at least 128 entries. Alternatively, a method to dynamically size the buffer should be implemented.=B4 If this is considered to be needed, I'll implement it. Thanks, Stefan -- M=FChlendamm 12 | Voice +49-40-256848, +49-177-3504009 D-22089 Hamburg | e-mail: stefan.bethke@hanse.de Germany | stb@freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 13:25:52 1999 Delivered-To: freebsd-net@freebsd.org Received: from mrdata.com (unknown [216.61.45.9]) by hub.freebsd.org (Postfix) with ESMTP id 5DC6E14EDE; Tue, 23 Mar 1999 13:24:23 -0800 (PST) (envelope-from blakef@mrdata.com) Received: (from blakef@localhost) by mrdata.com (8.8.8/8.8.5) id PAA03471; Tue, 23 Mar 1999 15:24:02 -0600 (CST) From: Blake Freeburg Message-Id: <199903232124.PAA03471@mrdata.com> Subject: Routing a subnet? To: freebsd-net@freebsd.org, freebsd-questions@freebsd.org Date: Tue, 23 Mar 1999 15:24:02 -0600 (CST) X-Mailer: ELM [version 2.4ME+ PL40 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello, I know this may be a simple routing problem, but I would like to know: How to route a subnet.... When we try to route route add -net 216.61.45.24/29 216.61.45.21 or route add 216.61.45.24/29 216.61.45.21 we always get a 'default' added to the routing tables which point to the 2nd router (216.61.45.21), not our default router (216.61.45.1). Am I missing something in the config settings? (both on 2.2.8 and 3.1?) Blake To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 13:30: 1 1999 Delivered-To: freebsd-net@freebsd.org Received: from mrdata.com (unknown [216.61.45.9]) by hub.freebsd.org (Postfix) with ESMTP id 4E2A214E58; Tue, 23 Mar 1999 13:28:38 -0800 (PST) (envelope-from blakef@mrdata.com) Received: (from blakef@localhost) by mrdata.com (8.8.8/8.8.5) id PAA03501; Tue, 23 Mar 1999 15:28:19 -0600 (CST) From: Blake Freeburg Message-Id: <199903232128.PAA03501@mrdata.com> Subject: Proper virtual domain setup? To: freebsd-questions@freebsd.org, freebsd-net@freebsd.org Date: Tue, 23 Mar 1999 15:28:19 -0600 (CST) X-Mailer: ELM [version 2.4ME+ PL40 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello, To configure a virtual domain with 2.2.8 or 3.1, I have been ifconfig .... alias route ... localhost now this seems to create a virtual alias, but it seems that my card is never seen when I do 'netstat -r'. Instead, the address set in rc.conf needs to be looped back to localhost. So I am wondering, am I doing this right? Any better way to do this? Blake To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 14: 1:17 1999 Delivered-To: freebsd-net@freebsd.org Received: from sims-ha.videotron.net (faure.videotron.net [205.151.222.100]) by hub.freebsd.org (Postfix) with ESMTP id 77D8B14EB3 for ; Tue, 23 Mar 1999 14:01:08 -0800 (PST) (envelope-from pcbroch@videotron.ca) Received: from shark ([207.96.177.9]) by sims-ha.videotron.net (Sun Internet Mail Server sims.3.5.1998.03.08.20.27) with SMTP id <0F92001IQHLA4L@sims-ha.videotron.net> for freebsd-net@freebsd.org; Tue, 23 Mar 1999 16:41:35 -0500 (EST) Date: Tue, 23 Mar 1999 16:44:54 -0500 From: Patrick Brochu Subject: Sharing a Cable modem connection To: freebsd-net@freebsd.org Message-id: <001d01be7576$635ecae0$0101a8c0@videotron.ca> MIME-version: 1.0 X-Mailer: Microsoft Outlook Express 5.00.2014.211 Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 7bit X-MSMail-Priority: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2014.211 X-Priority: 3 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi All: Very simple question, hope the answer is as simple! I have a cable modem connected to ed1, configured by DHCP I have my local network connected to ed2, configured with fixed ip address 192.168.1.1 Both interfaces work properly, i can surf the web as well as browse my local net. I would like to allow additional Win98 machines to access the internet through my FreeBSD pc. Anybody could recommend a good way of doing that? Ideally this would not require me to set up any proxies or things like that on my Win98 clients besides the Gateway. Thanks a lot! To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 14:11:16 1999 Delivered-To: freebsd-net@freebsd.org Received: from xylan.com (postal.xylan.com [208.8.0.248]) by hub.freebsd.org (Postfix) with ESMTP id CE47014CE8 for ; Tue, 23 Mar 1999 14:10:54 -0800 (PST) (envelope-from wes@softweyr.com) Received: from mailhub.xylan.com by xylan.com (8.8.7/SMI-SVR4 (xylan-mgw 2.2 [OUT])) id OAA19009; Tue, 23 Mar 1999 14:10:33 -0800 (PST) Received: from utah.XYLAN.COM by mailhub.xylan.com (SMI-8.6/SMI-SVR4 (mailhub 2.1 [HUB])) id OAA07943; Tue, 23 Mar 1999 14:10:32 -0800 Received: from softweyr.com by utah.XYLAN.COM (SMI-8.6/SMI-SVR4 (xylan utah [SPOOL])) id PAA15205; Tue, 23 Mar 1999 15:10:24 -0700 Message-ID: <36F8115D.85E79B47@softweyr.com> Date: Tue, 23 Mar 1999 15:10:37 -0700 From: Wes Peters Organization: Softweyr LLC X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Patrick Brochu Cc: freebsd-net@FreeBSD.ORG Subject: Re: Sharing a Cable modem connection References: <001d01be7576$635ecae0$0101a8c0@videotron.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Patrick Brochu wrote: > > Hi All: > Very simple question, hope the answer is as simple! > > I have a cable modem connected to ed1, configured by DHCP > I have my local network connected to ed2, configured with fixed ip address > 192.168.1.1 > > Both interfaces work properly, i can surf the web as well as browse my local > net. > > I would like to allow additional Win98 machines to access the internet > through my FreeBSD pc. Anybody could recommend a good way of doing that? > > Ideally this would not require me to set up any proxies or things like that > on my Win98 clients besides the Gateway. natd. Search for NAT or natd in the handbook, and feel free to direct any specific questions you have to freebsd-questions@freebsd.org. You've got questions? They've got answers! ;^) -- Where am I, and what am I doing in this handbasket? Wes Peters +1.801.915.2061 Softweyr LLC wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 15: 4:58 1999 Delivered-To: freebsd-net@freebsd.org Received: from dnai.com (dnai.com [207.181.194.98]) by hub.freebsd.org (Postfix) with ESMTP id 5802A14BD6 for ; Tue, 23 Mar 1999 15:04:56 -0800 (PST) (envelope-from miket@dnai.com) Received: from einstein (dnai-207-181-255-11.dialup.dnai.com [207.181.255.11]) by dnai.com (8.8.8/8.8.8) with SMTP id PAA05876 for ; Tue, 23 Mar 1999 15:04:36 -0800 (PST) Message-Id: <4.1.19990323134009.0099fe10@mail.dnai.com> X-Sender: miket@mail.dnai.com X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Tue, 23 Mar 1999 15:03:39 -0800 To: freebsd-net@freebsd.org From: Mike Thompson Subject: IP alias configuration for high availability Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello everyone, I'm looking for comments on the following proposed scheme for managing IP addresses between various servers for high availability. I am attempting to come up with network configuration where our web servers running FreeBSD 2.2.8 will respond to two different IP addresses on the same network interface. The idea is that first network address for a server would be fixed on the reserved IP addresses of 192.168.xxx.xxx (behind a router to prevent packet leakage onto the Internet). These fixed addresses would be used for private communication between the various servers. The other addresses used by each server would be from the Internet IP address block assigned from our ISP (example 209.185.152.1 thru 209.185.152.20). The Internet IP addresses would be round-robin load balanced from our DNS server. Under centralized control any server could be commanded to alias an IP address from this block. With such a configuration we hope to be able to hot swap servers among the various Internet IP addresses for system maintenance and emergency situations for high availability. To configure the fixed and dynamic addresses for a server I can use the following commands: ifconfig ed0 inet 192.168.0.1 netmask 255.255.255.0 ifconfig ed0 inet 209.185.152.1 netmask 255.255.255.0 alias To dynamically add another Internet IP address to the server I can then issue the command: ifconfig ed0 inet 209.185.152.2 netmask 255.255.255.0 alias To dynamically remove the IP address to the server I can issue the command: ifconfig ed0 inet 209.185.152.2 delete From my testing it seems everything works pretty well. I have found that I need to manually break the association between an IP address and ethernet address on the servers by using the "arp" utility. Are there other issues that I should be aware of by following such a strategy? Thanks, Mike Thompson To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 19:25:50 1999 Delivered-To: freebsd-net@freebsd.org Received: from phluffy.fks.bt (net25-cust199.pdx.wantweb.net [24.236.25.199]) by hub.freebsd.org (Postfix) with ESMTP id 064BB1537E for ; Tue, 23 Mar 1999 19:25:31 -0800 (PST) (envelope-from myke@ees.com) Received: from localhost (myke@localhost) by phluffy.fks.bt (8.8.8/8.8.8) with ESMTP id TAA14935; Tue, 23 Mar 1999 19:24:59 -0800 (PST) (envelope-from myke@ees.com) Date: Tue, 23 Mar 1999 19:24:59 -0800 (PST) From: Mike Holling X-Sender: myke@phluffy.fks.bt To: Patrick Brochu Cc: freebsd-net@FreeBSD.ORG Subject: Re: Sharing a Cable modem connection In-Reply-To: <001d01be7576$635ecae0$0101a8c0@videotron.ca> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > I would like to allow additional Win98 machines to access the internet > through my FreeBSD pc. Anybody could recommend a good way of doing that? I've done this many times using natd. The natd man page has a good step-by-step procedure near the end. - Mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 23:10: 3 1999 Delivered-To: freebsd-net@freebsd.org Received: from bubba.whistle.com (s205m7.whistle.com [207.76.205.7]) by hub.freebsd.org (Postfix) with ESMTP id E51AA1540E for ; Tue, 23 Mar 1999 23:08:38 -0800 (PST) (envelope-from archie@whistle.com) Received: (from archie@localhost) by bubba.whistle.com (8.9.2/8.9.2) id XAA53388; Tue, 23 Mar 1999 23:08:11 -0800 (PST) From: Archie Cobbs Message-Id: <199903240708.XAA53388@bubba.whistle.com> Subject: Re: Problem with mpd2.0b1 In-Reply-To: <004101be6cd3$ebcc8120$fd43c3d1@tytus.blackgate.com> from Wayne MacLaurin at "Mar 12, 99 05:01:46 pm" To: waynem@cyberus.ca (Wayne MacLaurin) Date: Tue, 23 Mar 1999 23:08:11 -0800 (PST) Cc: freebsd-net@FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Wayne MacLaurin writes: > We are having two separate problem with mpd2.0b1 running on FreeBSD 2.1.7. > We have two systems with four internal USR 33.6 Sportster modems. > > 1) If we build a bundle with more than two modems, when the systems connect > both ends simultaneously dump core (segv). You should first try upgrading your system to 2.2.8. FreeBSD 2.1.7 is quite old. Also, post some diagnostic info, ie., the log trace and (even better) a gdb stack trace (type "gdb mpd mpd.core" and then "where"). > 2) Not matter how many modems (1 or 2) we try to connect and no matter what > "Connect" speed is reported back by the chat script, the "bandwidth" report > is always 9600. Even when it does connect with two modems, the bandwidth is > always 9600. I've seen Connect messages of 14400, 21600 and various > others... The script reports what the modem reports (if you say "log +chat2" you can verify this).. however the "set link bandwidth" values are *not* automatically updated from the output of the chat script yet. I hope to fix this in the next version. -Archie ___________________________________________________________________________ Archie Cobbs * Whistle Communications, Inc. * http://www.whistle.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 23:31:24 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 4CB3B14E42 for ; Tue, 23 Mar 1999 23:31:16 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id IAA00921 for ; Wed, 24 Mar 1999 08:30:56 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Problems and Questions to bridge Date: Wed, 24 Mar 1999 08:20:18 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99032408331102.00328@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, I have the following scenario: PC1:192.168.0.238 PC2:192.168.0.96 |---------| Bridge, FBSD3.1 |---------| | | |-------------------| | | | xl0|------|ed1 ed0|-------| | | | |-------------------| | | |---------| |---------| I configured the bridge as described in the manual: sysctl -w net.link.ether.bridge=1 sysctl -w net.link.ether.bridge_ipfw=1 Problem: I can't ping from PC1 to PC2 or from PC2 to PC1. If I assign an IP-address to the bridge, I can ping from PC1 to the bridge and from PC2 to the bridge, but not through the bridge. I use also ipfw on the bridge and here is the output: 00150 0 0 allow ip from any to any via lo0 00200 0 0 deny ip from any to 127.0.0.0/8 00300 87 10385 allow ip from any to any in recv ed0 00400 0 0 allow ip from any to any out xmit ed0 00500 54 4536 allow ip from any to any in recv ed1 00600 59 4908 allow ip from any to any out xmit ed1 65535 0 0 allow ip from any to any It seems that no packet go out through ed0! a) What I've done wrong? Thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Tue Mar 23 23:33:33 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 7082314D06 for ; Tue, 23 Mar 1999 23:33:30 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id GAA09466; Wed, 24 Mar 1999 06:18:41 +0100 From: Luigi Rizzo Message-Id: <199903240518.GAA09466@labinfo.iet.unipi.it> Subject: Re: Problems and Questions to bridge To: mbretter@sbox.tu-graz.ac.at (Michael Bretterklieber) Date: Wed, 24 Mar 1999 06:18:40 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: <99032408331102.00328@babsi.tu-graz.ac.at> from "Michael Bretterklieber" at Mar 24, 99 08:19:59 am X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1962 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Hi, > > I have the following scenario: > > PC1:192.168.0.238 PC2:192.168.0.96 > |---------| Bridge, FBSD3.1 |---------| > | | |-------------------| | | > | xl0|------|ed1 ed0|-------| | > | | |-------------------| | | > |---------| |---------| you need 3.1-stable, or at least patch /sys/i386/isa/if_ed.c to what is in -stable. By mistake i forgot to put in some pieces when i ported the bridging code to the 3.x branch and realized that only recently. cheers luigi > I configured the bridge as described in the manual: > sysctl -w net.link.ether.bridge=1 > sysctl -w net.link.ether.bridge_ipfw=1 > > Problem: I can't ping from PC1 to PC2 or from PC2 to PC1. > If I assign an IP-address to the bridge, > I can ping from PC1 to the bridge and from PC2 to the > bridge, but not through the bridge. > I use also ipfw on the bridge and here is the output: > > 00150 0 0 allow ip from any to any via lo0 > 00200 0 0 deny ip from any to 127.0.0.0/8 > 00300 87 10385 allow ip from any to any in recv ed0 > 00400 0 0 allow ip from any to any out xmit ed0 > 00500 54 4536 allow ip from any to any in recv ed1 > 00600 59 4908 allow ip from any to any out xmit ed1 > 65535 0 0 allow ip from any to any > > It seems that no packet go out through ed0! > > a) What I've done wrong? > > Thanx, > -- > \|/ > @ @ > +---------------------------oOO-(_)-OOo---------------------------+ > Michael Bretterklieber > Office: Michael.Bretterklieber@gamed.com > Privat: mbretter@sbox.tu-graz.ac.at > URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ > +-----------------------------------------------------------------+ > > > 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 Wed Mar 24 2: 7:15 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 0FA2F14BE9 for ; Wed, 24 Mar 1999 02:07:07 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id LAA05682 for ; Wed, 24 Mar 1999 11:06:47 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Bridge and dummynet Date: Wed, 24 Mar 1999 10:52:10 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99032411090200.05750@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, I have the following scenario: PC1:192.168.0.238 PC2:192.168.0.96 |---------| Bridge, FBSD3.1 |---------| | | |-------------------| | | | xl0|------|ed1 ed0|-------| | | | |-------------------| | | |---------| |---------| The bridge works with the patch /sys/i386/isa/if_ed.c Version 1.150. Now I'd like to limit the transferrate down to 64KBit/s between PC1 and PC2. Here are my rules: 00020 0 0 pipe 20 ip from 192.168.0.238 to any out xmit ed0 00021 8 712 pipe 21 ip from 192.168.0.238 to any out xmit ed1 00022 0 0 pipe 22 ip from any to 192.168.0.238 out xmit ed1 00023 11 600 pipe 23 ip from any to 192.168.0.238 out xmit ed0 00150 0 0 allow ip from any to any via lo0 00200 0 0 deny ip from any to 127.0.0.0/8 00300 405 23772 allow ip from any to any out xmit ed0 00400 56 6149 allow ip from any to any in recv ed0 00500 262 42583 allow ip from any to any out xmit ed1 00600 73 5263 allow ip from any to any in recv ed1 65535 31 0 allow ip from any to any I configuered the pipes with: ipfw pipe 20 config bw 64KBits/s ipfw pipe 21 config bw 64KBits/s ipfw pipe 22 config bw 64KBits/s ipfw pipe 23 config bw 64KBits/s Result: The connect between PC1 and PC2 broke up, on the bridge I got the kernel message: "dummynet: bad switch 3!" I deleted the pipes and everything was ok. I was wondering why no packets passed rule 20 and 22. The firewall-verbose-logformat changed between 3.0 and 3.1, no ports are displayed, only source and destination-ip are printed out?! Thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 2:21:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id BF635150F6 for ; Wed, 24 Mar 1999 02:19:29 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id JAA09727; Wed, 24 Mar 1999 09:04:36 +0100 From: Luigi Rizzo Message-Id: <199903240804.JAA09727@labinfo.iet.unipi.it> Subject: Re: Bridge and dummynet To: mbretter@sbox.tu-graz.ac.at (Michael Bretterklieber) Date: Wed, 24 Mar 1999 09:04:36 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: <99032411090200.05750@babsi.tu-graz.ac.at> from "Michael Bretterklieber" at Mar 24, 99 10:51:51 am X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 2920 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Hi, > > I have the following scenario: > > PC1:192.168.0.238 PC2:192.168.0.96 > |---------| Bridge, FBSD3.1 |---------| > | | |-------------------| | | > | xl0|------|ed1 ed0|-------| | > | | |-------------------| | | > |---------| |---------| > > The bridge works with the patch /sys/i386/isa/if_ed.c Version 1.150. > Now I'd like to limit the transferrate down to 64KBit/s between PC1 and PC2. > Here are my rules: > > 00020 0 0 pipe 20 ip from 192.168.0.238 to any out xmit ed0 > 00021 8 712 pipe 21 ip from 192.168.0.238 to any out xmit ed1 > 00022 0 0 pipe 22 ip from any to 192.168.0.238 out xmit ed1 > 00023 11 600 pipe 23 ip from any to 192.168.0.238 out xmit ed0 > 00150 0 0 allow ip from any to any via lo0 > 00200 0 0 deny ip from any to 127.0.0.0/8 > 00300 405 23772 allow ip from any to any out xmit ed0 > 00400 56 6149 allow ip from any to any in recv ed0 > 00500 262 42583 allow ip from any to any out xmit ed1 > 00600 73 5263 allow ip from any to any in recv ed1 > 65535 31 0 allow ip from any to any > > I configuered the pipes with: > ipfw pipe 20 config bw 64KBits/s > ipfw pipe 21 config bw 64KBits/s > ipfw pipe 22 config bw 64KBits/s > ipfw pipe 23 config bw 64KBits/s > > Result: > The connect between PC1 and PC2 broke up, on the bridge I got > the kernel message: "dummynet: bad switch 3!" hmmmm there is probably a missing #include "opt_bdg.h" in the beginning of ip_dummynet.c near the #ifdef BRIDGE section > I deleted the pipes and everything was ok. > > I was wondering why no packets passed rule 20 and 22. two reasons: 1) you also have to update sys/net/bridge.c and sys/netinet/ip_fw.c and (probably) sys/netinet/ip_dummynet.c 2) the 'out' specifier in ipfw rules does not work with briding cannot work because the bridging code does not know which output interface will be used (it could be sent to all). You can only do selection basing on input interface or better just on addresses -- the firewall will act only once on bridged packets instead of the two times with routed pkts. cheers luigi > The firewall-verbose-logformat changed between 3.0 and 3.1, > no ports are displayed, only source and destination-ip are > printed out?! > > Thanx, > -- > \|/ > @ @ > +---------------------------oOO-(_)-OOo---------------------------+ > Michael Bretterklieber > Office: Michael.Bretterklieber@gamed.com > Privat: mbretter@sbox.tu-graz.ac.at > URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ > +-----------------------------------------------------------------+ > > > 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 Wed Mar 24 4:14:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 158641526B for ; Wed, 24 Mar 1999 04:14:17 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id NAA08638 for ; Wed, 24 Mar 1999 13:13:46 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Re: bridge and dummynet Date: Wed, 24 Mar 1999 13:12:08 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99032413160200.12899@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, Now dummynet with bridge works, I updatet the following files: /sys/net/bridge.c /sys/netinet/ip_fw.c /sys/netinet/ip_dummynet.c before "#ifdef DRIDGE", I added "#include "opt_bdg.h" /sys/i386/isa/if_ed.c Thanx, Thanx Luigi, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 8: 1:40 1999 Delivered-To: freebsd-net@freebsd.org Received: from freelove.tu-graz.ac.at (freelove.tu-graz.ac.at [129.27.193.131]) by hub.freebsd.org (Postfix) with SMTP id 1454914CC4 for ; Wed, 24 Mar 1999 08:01:01 -0800 (PST) (envelope-from mbretter@freelove.tu-graz.ac.at) Received: (from mbretter@localhost) by freelove.tu-graz.ac.at (8.6.11/8.6.9) id RAA08508; Wed, 24 Mar 1999 17:00:42 +0100 Date: Wed, 24 Mar 1999 17:00:41 +0100 (MET) From: mbretter X-Sender: mbretter@freelove.tu-graz.ac.at To: freebsd-net@freebsd.org Subject: Dummynet 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, I have a new problem with bandwidth limiting: I added: pipe 1 ip from A to B and a bw of 64KBytes/s results in an effective bw of 2.5KBytes/s (counted via ftp) 128KBytes/s -> 5 KBytes/s 256KBytes/s -> 10 KBytes/s etc...?! Thanx, MICHAEL BRETTERKLIEBER alias mbretter@sbox.tu-graz.ac.at To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 8:16:15 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 8AC1514C9B for ; Wed, 24 Mar 1999 08:16:09 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id PAA10220; Wed, 24 Mar 1999 15:01:19 +0100 From: Luigi Rizzo Message-Id: <199903241401.PAA10220@labinfo.iet.unipi.it> Subject: Re: Dummynet To: mbretter@sbox.tu-graz.ac.at (mbretter) Date: Wed, 24 Mar 1999 15:01:18 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: from "mbretter" at Mar 24, 99 05:00:22 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 451 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Hi, > > I have a new problem with bandwidth limiting: > > I added: pipe 1 ip from A to B > > and a bw of 64KBytes/s results in an effective bw of 2.5KBytes/s > (counted via ftp) > 128KBytes/s -> 5 KBytes/s > 256KBytes/s -> 10 KBytes/s > etc...?! send me the output of "ipfw show" and "ipfw pipe show" plus how big is the file -- for a very short one you won't be able to measure the real throughput using ftp because of tcp slow start. luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 9:27: 8 1999 Delivered-To: freebsd-net@freebsd.org Received: from alpha.netvision.net.il (alpha.netvision.net.il [194.90.1.13]) by hub.freebsd.org (Postfix) with ESMTP id 9457E154B5 for ; Wed, 24 Mar 1999 09:26:53 -0800 (PST) (envelope-from sharon_j@netvision.net.il) Received: from netvision.net.il (RAS7-p50.hfa.netvision.net.il [62.0.149.50]) by alpha.netvision.net.il (8.9.3/8.8.6) with ESMTP id TAA08435 for ; Wed, 24 Mar 1999 19:26:32 +0200 (IST) Message-ID: <36F91ED2.C42E686C@netvision.net.il> Date: Wed, 24 Mar 1999 19:20:18 +0200 From: Chen Genossar X-Mailer: Mozilla 4.05 [en] (Win95; I) MIME-Version: 1.0 To: freebsd-net@FreeBSD.ORG Subject: configuration tools for gated Content-Type: text/plain; charset=x-user-defined Content-Transfer-Encoding: 8bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, I want to use gated as a router. I have some questions about monitoring & administrating tools for gated: I know there is two tools: rip-query & ospf-monitor. Those tools are very static (can't make dynamic changes with them). I am looking for dynamic configuration tools. Does anybody knows which administrative tools exists for gated? Do they (this tools) are working with CLI ? GUI ? Does this tools support dynamic modifications & configuration? Can anybody recommend me on a tool / tools for dynamic configuration of routers ?  Thanks in advance for any help     Sharon.   To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 9:45:32 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 748D014D97 for ; Wed, 24 Mar 1999 09:45:01 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id SAA14733 for ; Wed, 24 Mar 1999 18:44:41 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Re: Dummynet Date: Wed, 24 Mar 1999 18:33:27 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain References: <199903241401.PAA10220@labinfo.iet.unipi.it> MIME-Version: 1.0 Message-Id: <99032418465600.02469@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, > send me the output of "ipfw show" and "ipfw pipe show" > plus how big is the file -- for a very short one you won't be able to > measure the real throughput using ftp because of tcp slow start. > > luigi I tried it with two PC's without any problems: PC1 Bridge |----| 64kBit |---------| | xl0|----------|ed1 ed0|-----> Nirvana, no IP-Addr. |----| |---------| 192.168.0.238 192.168.0.1 The configuartion with problems: PC1 Bridge PC2 |----| 64kBit |---------| |------| | xl0|----------|ed1 ed0|-----| | |----| |---------| |------| 192.168.0.238 no IP 192.168.0.96 Mybe the problem occurs, because I have no IP-Addr. assigned to the bridge? Tomorrow I'll send you the output of ipfw show and ipfw pipe show, because I'm now at home, where I have only two PC's Thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 11:31:37 1999 Delivered-To: freebsd-net@freebsd.org Received: from lms1.cyber1.net (lms1.cyber1.net [208.206.222.92]) by hub.freebsd.org (Postfix) with ESMTP id B21E114DC3 for ; Wed, 24 Mar 1999 11:31:34 -0800 (PST) (envelope-from peter@cyber1.net) Received: from localhost (peter@localhost) by lms1.cyber1.net (8.8.7/8.8.7) with SMTP id OAA10074 for ; Wed, 24 Mar 1999 14:31:24 -0500 Date: Wed, 24 Mar 1999 14:31:23 -0500 (EST) From: Peter Brezny To: freebsd-net@freebsd.org Subject: switch vs bridge (fwd) 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 ---------- Forwarded message ---------- Date: Wed, 24 Mar 1999 14:02:50 -0500 (EST) From: Peter Brezny To: freebsd-isp@FreeBSD.ORG Subject: switch vs bridge is a switch the same thing as a multi port bridge? thanks. Peter Brezny cyber1.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" 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 Wed Mar 24 11:58:57 1999 Delivered-To: freebsd-net@freebsd.org Received: from hp9000.chc-chimes.com (hp9000.chc-chimes.com [206.67.97.84]) by hub.freebsd.org (Postfix) with ESMTP id 141EA14DB3 for ; Wed, 24 Mar 1999 11:58:14 -0800 (PST) (envelope-from billf@chc-chimes.com) Received: from localhost by hp9000.chc-chimes.com with SMTP (1.39.111.2/16.2) id AA180403381; Wed, 24 Mar 1999 14:23:01 -0500 Date: Wed, 24 Mar 1999 14:23:01 -0500 (EST) From: Bill Fumerola To: Peter Brezny Cc: freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) 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 Wed, 24 Mar 1999, Peter Brezny wrote: > is a switch the same thing as a multi port bridge? Not hardly. A switch replicates packets based on their Ethernet destination. - 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 Wed Mar 24 12:39:32 1999 Delivered-To: freebsd-net@freebsd.org Received: from silver.sms.fi (silver.sms.fi [194.111.122.17]) by hub.freebsd.org (Postfix) with SMTP id 4457514CEF for ; Wed, 24 Mar 1999 12:39:28 -0800 (PST) (envelope-from pete@sms.fi) Received: from sms.fi (localhost.sms.fi [127.0.0.1]) by silver.sms.fi (8.9.2/8.9.2) with ESMTP id WAA17481; Wed, 24 Mar 1999 22:38:26 +0200 (EET) (envelope-from pete@sms.fi) Message-ID: <36F94D41.FBEC5739@sms.fi> Date: Wed, 24 Mar 1999 22:38:25 +0200 From: Petri Helenius X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 3.1-STABLE i386) X-Accept-Language: en,fi MIME-Version: 1.0 To: Bill Fumerola Cc: Peter Brezny , freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Bill Fumerola wrote: > > On Wed, 24 Mar 1999, Peter Brezny wrote: > > > is a switch the same thing as a multi port bridge? > > Not hardly. > > A switch replicates packets based on their Ethernet destination. > And how exactly would you describe a bridge then? Pete To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 12:45:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from hp9000.chc-chimes.com (hp9000.chc-chimes.com [206.67.97.84]) by hub.freebsd.org (Postfix) with ESMTP id 4A45B14BCF for ; Wed, 24 Mar 1999 12:45:42 -0800 (PST) (envelope-from billf@chc-chimes.com) Received: from localhost by hp9000.chc-chimes.com with SMTP (1.39.111.2/16.2) id AA197996232; Wed, 24 Mar 1999 15:10:32 -0500 Date: Wed, 24 Mar 1999 15:10:32 -0500 (EST) From: Bill Fumerola To: Petri Helenius Cc: Peter Brezny , freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) In-Reply-To: <36F94D41.FBEC5739@sms.fi> 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 Wed, 24 Mar 1999, Petri Helenius wrote: > > A switch replicates packets based on their Ethernet destination. > > > And how exactly would you describe a bridge then? Anything I say is going to be able to be picked apart, but a bridge replicates any data it sees for its destination network and just shoves it over. A switch makes an intelligent port-by-port decision. - 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 Wed Mar 24 12:52:41 1999 Delivered-To: freebsd-net@freebsd.org Received: from silver.sms.fi (silver.sms.fi [194.111.122.17]) by hub.freebsd.org (Postfix) with SMTP id 8111E14F75 for ; Wed, 24 Mar 1999 12:52:38 -0800 (PST) (envelope-from pete@sms.fi) Received: from sms.fi (localhost.sms.fi [127.0.0.1]) by silver.sms.fi (8.9.2/8.9.2) with ESMTP id WAA17531; Wed, 24 Mar 1999 22:51:49 +0200 (EET) (envelope-from pete@sms.fi) Message-ID: <36F95064.670D0DA1@sms.fi> Date: Wed, 24 Mar 1999 22:51:48 +0200 From: Petri Helenius X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 3.1-STABLE i386) X-Accept-Language: en,fi MIME-Version: 1.0 To: Bill Fumerola Cc: Peter Brezny , freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Bill Fumerola wrote: > > On Wed, 24 Mar 1999, Petri Helenius wrote: > > > > A switch replicates packets based on their Ethernet destination. > > > > > And how exactly would you describe a bridge then? > > Anything I say is going to be able to be picked apart, > > but a bridge replicates any data it sees for its destination network and > just shoves it over. > > A switch makes an intelligent port-by-port decision. > 99% of ethernet bridges out there have a forwarding table as the switches have which are multiport bridges. It's a completely different ballgame for token ring though. There is no concept of "network" in ethernet so you cannot do any actions based on the addressing other than learn where all the sources are. This process is equally the same for bridges and switches. (since they are the same) Pete To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 13: 9:20 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 5D09615147 for ; Wed, 24 Mar 1999 13:09:10 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id TAA11277; Wed, 24 Mar 1999 19:54:00 +0100 From: Luigi Rizzo Message-Id: <199903241854.TAA11277@labinfo.iet.unipi.it> Subject: Re: switch vs bridge (fwd) To: peter@cyber1.net (Peter Brezny) Date: Wed, 24 Mar 1999 19:54:00 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: from "Peter Brezny" at Mar 24, 99 02:31:04 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 67 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > is a switch the same thing as a multi port bridge? yes. luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 13:14:43 1999 Delivered-To: freebsd-net@freebsd.org Received: from xylan.com (postal.xylan.com [208.8.0.248]) by hub.freebsd.org (Postfix) with ESMTP id E955A14FAE for ; Wed, 24 Mar 1999 13:14:39 -0800 (PST) (envelope-from wes@softweyr.com) Received: from mailhub.xylan.com by xylan.com (8.8.7/SMI-SVR4 (xylan-mgw 2.2 [OUT])) id NAA00805; Wed, 24 Mar 1999 13:12:31 -0800 (PST) Received: from utah.XYLAN.COM by mailhub.xylan.com (SMI-8.6/SMI-SVR4 (mailhub 2.1 [HUB])) id NAA18492; Wed, 24 Mar 1999 13:12:30 -0800 Received: from softweyr.com by utah.XYLAN.COM (SMI-8.6/SMI-SVR4 (xylan utah [SPOOL])) id OAA07619; Wed, 24 Mar 1999 14:12:21 -0700 Message-ID: <36F95545.6B1F40C0@softweyr.com> Date: Wed, 24 Mar 1999 14:12:37 -0700 From: Wes Peters Organization: Softweyr LLC X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Petri Helenius Cc: Bill Fumerola , Peter Brezny , freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) References: <36F95064.670D0DA1@sms.fi> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Petri Helenius wrote: > > Bill Fumerola wrote: > > > > On Wed, 24 Mar 1999, Petri Helenius wrote: > > > > > > A switch replicates packets based on their Ethernet destination. > > > > > > > And how exactly would you describe a bridge then? > > > > Anything I say is going to be able to be picked apart, > > > > but a bridge replicates any data it sees for its destination network and > > just shoves it over. > > > > A switch makes an intelligent port-by-port decision. > > > 99% of ethernet bridges out there have a forwarding table as the switches > have which are multiport bridges. It's a completely different ballgame for > token ring though. > > There is no concept of "network" in ethernet so you cannot do any actions > based on the addressing other than learn where all the sources are. This > process > is equally the same for bridges and switches. (since they are the same) Try these on for size: MAC-Layer Bridge - A device used to forward data between LANs at layer two, by automatically filtering out traffic which is local to each LAN, while forwarding on traffic which is not local to each LAN. All broadcasts and multicasts, as well as all traffic with a destination address which has not been learned by the bridge, is forwarded. MAC-Layer Switching - LAN data transferred through a network based on the source and destination address contained in the MAC header of the frame. Essentially the same as bridging, but almost always employing dedicated hardware to perform the switching. In other words, there isn't a whole lot of difference unless you step up to layer three switches. These definitions taken from "The Switching Book II", from Xylan Corp. Download a PDF version or order your own free printed copy at: http://www.xylan.com/library/switchbook/index.html While you're there, poke around and buy something. ;^) -- Where am I, and what am I doing in this handbasket? Wes Peters +1.801.915.2061 Softweyr LLC wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 13:23:43 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 589EE1503D for ; Wed, 24 Mar 1999 13:23:37 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id UAA11367; Wed, 24 Mar 1999 20:08:15 +0100 From: Luigi Rizzo Message-Id: <199903241908.UAA11367@labinfo.iet.unipi.it> Subject: Re: switch vs bridge (fwd) To: wes@softweyr.com (Wes Peters) Date: Wed, 24 Mar 1999 20:08:14 +0100 (MET) Cc: pete@sms.fi, billf@chc-chimes.com, peter@cyber1.net, freebsd-net@FreeBSD.ORG In-Reply-To: <36F95545.6B1F40C0@softweyr.com> from "Wes Peters" at Mar 24, 99 02:12:18 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1413 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > is equally the same for bridges and switches. (since they are the same) > > Try these on for size: it seems to me that in the definitions below "switching" is the action and "bridge" is the device performing it. > MAC-Layer Bridge - A device used to forward data between LANs at layer two, > by automatically filtering out traffic which is local to each LAN, while > forwarding on traffic which is not local to each LAN. All broadcasts > and multicasts, as well as all traffic with a destination address which > has not been learned by the bridge, is forwarded. > > MAC-Layer Switching - LAN data transferred through a network based on the > source and destination address contained in the MAC header of the frame. > Essentially the same as bridging, but almost always employing dedicated > hardware to perform the switching. quite a few commercial bridges do use "dedicated hardware", only hackers use software based solutions (PCBRIDGE in the early days, linux/freebsd based solutions these days...) cheers luigi -----------------------------------+------------------------------------- Luigi RIZZO . EMAIL: luigi@iet.unipi.it . Dip. di Ing. dell'Informazione HTTP://www.iet.unipi.it/~luigi/ . Universita` di Pisa TEL/FAX: +39-050-568.533/522 . via Diotisalvi 2, 56126 PISA (Italy) -----------------------------------+------------------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 13:24:45 1999 Delivered-To: freebsd-net@freebsd.org Received: from sumatra.americantv.com (sumatra.americantv.com [207.170.17.37]) by hub.freebsd.org (Postfix) with ESMTP id 461BC14EDA for ; Wed, 24 Mar 1999 13:24:39 -0800 (PST) (envelope-from jlemon@americantv.com) Received: from right.PCS (right.PCS [148.105.10.31]) by sumatra.americantv.com (8.8.5/8.8.5) with ESMTP id PAA17466; Wed, 24 Mar 1999 15:24:20 -0600 (CST) Received: from free.pcs (free.PCS [148.105.10.51]) by right.PCS (8.6.13/8.6.4) with ESMTP id PAA13137; Wed, 24 Mar 1999 15:23:48 -0600 Received: (from jlemon@localhost) by free.pcs (8.8.6/8.8.5) id PAA11962; Wed, 24 Mar 1999 15:23:48 -0600 (CST) Date: Wed, 24 Mar 1999 15:23:48 -0600 (CST) From: Jonathan Lemon Message-Id: <199903242123.PAA11962@free.pcs> To: wes@softweyr.com, freebsd-net@freebsd.org Subject: Re: switch vs bridge (fwd) X-Newsgroups: local.mail.freebsd-net In-Reply-To: References: Organization: Architecture and Operating System Fanatics Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In article you write: >In other words, there isn't a whole lot of difference unless you step >up to layer three switches. > >These definitions taken from "The Switching Book II", from Xylan Corp. >Download a PDF version or order your own free printed copy at: > > http://www.xylan.com/library/switchbook/index.html So where's the definition for layer-four switches? (What the heck is a layer 4 switch anyway?) -- Jonathan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 13:47:10 1999 Delivered-To: freebsd-net@freebsd.org Received: from lms1.cyber1.net (lms1.cyber1.net [208.206.222.92]) by hub.freebsd.org (Postfix) with ESMTP id 3092114C0B for ; Wed, 24 Mar 1999 13:47:06 -0800 (PST) (envelope-from peter@cyber1.net) Received: from localhost (peter@localhost) by lms1.cyber1.net (8.8.7/8.8.7) with SMTP id QAA11009; Wed, 24 Mar 1999 16:46:31 -0500 Date: Wed, 24 Mar 1999 16:46:31 -0500 (EST) From: Peter Brezny To: Jonathan Lemon Cc: wes@softweyr.com, freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) In-Reply-To: <199903242123.PAA11962@free.pcs> 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 layer 4... my friend keith's understanding is that layer 4 switches examine actual content of ip frames. usefull for multimedia quality of service and bandwidth reservation. sort of gives atm style qos and managability to ethernet. hope that helps. pb On Wed, 24 Mar 1999, Jonathan Lemon wrote: > In article you write: > >In other words, there isn't a whole lot of difference unless you step > >up to layer three switches. > > > >These definitions taken from "The Switching Book II", from Xylan Corp. > >Download a PDF version or order your own free printed copy at: > > > > http://www.xylan.com/library/switchbook/index.html > > So where's the definition for layer-four switches? (What the heck > is a layer 4 switch anyway?) > -- > Jonathan > > > 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 Wed Mar 24 17:11: 3 1999 Delivered-To: freebsd-net@freebsd.org Received: from xylan.com (postal.xylan.com [208.8.0.248]) by hub.freebsd.org (Postfix) with ESMTP id 5757314FD8 for ; Wed, 24 Mar 1999 17:10:56 -0800 (PST) (envelope-from wes@softweyr.com) Received: from mailhub.xylan.com by xylan.com (8.8.7/SMI-SVR4 (xylan-mgw 2.2 [OUT])) id RAA03423; Wed, 24 Mar 1999 17:10:36 -0800 (PST) Received: from utah.XYLAN.COM by mailhub.xylan.com (SMI-8.6/SMI-SVR4 (mailhub 2.1 [HUB])) id RAA27042; Wed, 24 Mar 1999 17:10:35 -0800 Received: from thrallo.utah.xylan.com by utah.XYLAN.COM (SMI-8.6/SMI-SVR4 (xylan utah [SPOOL])) id SAA22324; Wed, 24 Mar 1999 18:10:25 -0700 Subject: Re: switch vs bridge (fwd) From: Wes Peters Reply-To: Wes Peters In-Reply-To: <199903242123.PAA11962@free.pcs> Message-ID: <000346da2d1cdc6c_mailit@thrallo.utah.xylan.com> References: <199903242123.PAA11962@free.pcs> Date: Wed, 24 Mar 1999 18:10:03 -0800 X-Mailer: BeatWare Mail-It 2.0 X-BeOS-Platform: Intel or clone X-Priority: 3 (Normal) To: freebsd-net@freebsd.org, jlemon@americantv.com Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >In article you write: >>In other words, there isn't a whole lot of difference unless you step >>up to layer three switches. >> >>These definitions taken from "The Switching Book II", from Xylan Corp. >>Download a PDF version or order your own free printed copy at: >> >> http://www.xylan.com/library/switchbook/index.html > >So where's the definition for layer-four switches? (What the heck >is a layer 4 switch anyway?) Layer 4 switching is a marketing term used by those who don't have the money or technical ability to implement REAL QoS (Quality of Service) in their switches. ;^) The idea behind layer 4 switching is to look into the packet contents and be able to do semi-intelligent things based on that information, like directing HTTP streams to multiple servers. It's great for those few instances where you really need that, like high-volume web servers, and a complete boondoggle for most of the networking world. But it sounds good, doesn't it? If layer 2 switching is good, and layer 3 switching is great, then layer 4 switching must be like getting laid while eating your favorite ice cream, right? Sorry, I'm in a nasty mood right now, and using an unfamiliar mailer to boot. I'll be nice again tomorrow. -- Wes To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Wed Mar 24 17:34:11 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 144C814EB8 for ; Wed, 24 Mar 1999 17:34:05 -0800 (PST) (envelope-from mike@sentex.net) Received: from ospf-wat.sentex.net (ospf-wat.sentex.net [209.167.248.81]) by granite.sentex.net (8.8.8/8.6.9) with SMTP id UAA03265; Wed, 24 Mar 1999 20:33:40 -0500 (EST) From: mike@sentex.net (Mike Tancsa) To: sharon_j@netvision.net.il (Chen Genossar) Cc: freebsd-net@FreeBSD.ORG Subject: Re: configuration tools for gated Date: Thu, 25 Mar 1999 01:43:13 GMT Message-ID: <36f98643.3528635239@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 24 Mar 1999 12:28:25 -0500, in sentex.lists.freebsd.misc you wrote: >Hi, > >I want to use gated as a router. >I have some questions about monitoring & administrating tools for gated: > >I know there is two tools: rip-query & ospf-monitor. >Those tools are very static (can't make dynamic changes with them). >I am looking for dynamic configuration tools. gdc checkconf gdc reconfig These will be your friends with gated. >Does anybody knows which administrative tools exists for gated? gdc checkconf gdc reconfig >Do they (this tools) are working with CLI ? GUI ? gdc checkconf gdc reconfig All from the shell prompt.. >Does this tools support dynamic modifications & configuration? Yes. Make sure you turn up the tracing/logging. It can be quite verbose and chatty, but it should give you most everything you need. Also, look to www.gated.org and the mailing list archives for more info. Other projects that might be of interest are www.zebra.org and mrt http://www.merit.edu/net-research/mrt/html/ gated is pretty sparse for documentation, and features. But it seems to work... OK. ---Mike Mike Tancsa (mdtancsa@sentex.net) Sentex Communications Corp, Waterloo, Ontario, Canada To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 7:42:33 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id EA4EF15391 for ; Thu, 25 Mar 1999 07:42:19 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id QAA01221 for ; Thu, 25 Mar 1999 16:41:50 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Re: Dummynet Date: Thu, 25 Mar 1999 16:21:29 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain References: <199903241401.PAA10220@labinfo.iet.unipi.it> MIME-Version: 1.0 Message-Id: <99032516440600.00307@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Am Mi , 24 Mär 1999 schrieben Sie: > > Hi, > > > > I have a new problem with bandwidth limiting: > > > > I added: pipe 1 ip from A to B > > > > and a bw of 64KBytes/s results in an effective bw of 2.5KBytes/s > > (counted via ftp) > > 128KBytes/s -> 5 KBytes/s > > 256KBytes/s -> 10 KBytes/s > > etc...?! > > send me the output of "ipfw show" and "ipfw pipe show" > plus how big is the file -- for a very short one you won't be able to > measure the real throughput using ftp because of tcp slow start. > > luigi PC1 Bridge PC2 |----| |---------| |------| | xl0|----------|ed1 ed0|-----| | |----| |---------| |------| 192.168.0.238 no IP 192.168.0.96 I transmitted one file with 56940 Bytes from PC1 to PC2 via ftp in 13.24 sec. (4.30 KByte/s): Here is the output from "ipfw show": 00010 69 75877 pipe 1 ip from 192.168.0.238 to 192.168.0.96 00150 0 0 allow ip from any to any via lo0 00200 0 0 deny ip from any to 127.0.0.0/8 65535 12 0 allow ip from any to any Here is the output from "ipfw pipe show": 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops Thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 9:47:23 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 2BF1814BCF for ; Thu, 25 Mar 1999 09:47:20 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id OAA12974; Thu, 25 Mar 1999 14:51:32 +0100 From: Luigi Rizzo Message-Id: <199903251351.OAA12974@labinfo.iet.unipi.it> Subject: Re: Dummynet To: mbretter@sbox.tu-graz.ac.at (Michael Bretterklieber) Date: Thu, 25 Mar 1999 14:51:32 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: <99032516440600.00307@babsi.tu-graz.ac.at> from "Michael Bretterklieber" at Mar 25, 99 04:21:10 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1566 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > > I added: pipe 1 ip from A to B > > > > > > and a bw of 64KBytes/s results in an effective bw of 2.5KBytes/s > > > (counted via ftp) > > > 128KBytes/s -> 5 KBytes/s > > > 256KBytes/s -> 10 KBytes/s > > > etc...?! > > > > send me the output of "ipfw show" and "ipfw pipe show" > > plus how big is the file -- for a very short one you won't be able to > > measure the real throughput using ftp because of tcp slow start. > > > > luigi > PC1 Bridge PC2 > |----| |---------| |------| > | xl0|----------|ed1 ed0|-----| | > |----| |---------| |------| > 192.168.0.238 no IP 192.168.0.96 > > I transmitted one file with 56940 Bytes from PC1 to PC2 via ftp > in 13.24 sec. (4.30 KByte/s): > > Here is the output from "ipfw show": > 00010 69 75877 pipe 1 ip from 192.168.0.238 to 192.168.0.96 > 00150 0 0 allow ip from any to any via lo0 > 00200 0 0 deny ip from any to 127.0.0.0/8 > 65535 12 0 allow ip from any to any > > Here is the output from "ipfw pipe show": > 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops there must be some problem somewhere, are you sure you have the latest bridge.c ipfw.c and ip_dummynet.c and binaries are derived from them -- it makes no sense that you have 10 pkts in the queue and certainly they cannot sum up to over 300Kbytes. (it might well be my fault... can you refresh me on your os and the version of the above files... in doubt rm kernel bridge.o ipfw.o ip_dummynet.o from your compile dir and rebuild one.) luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13: 3: 2 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 8DF4015576; Thu, 25 Mar 1999 13:02:52 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id NAA00571; Thu, 25 Mar 1999 13:02:33 -0800 (PST) Date: Thu, 25 Mar 1999 13:02:33 -0800 (PST) From: Jesse To: freebsd-net@freebsd.org, freebsd-hackers@freebsd.org Subject: 3.1-STABLE dies on 40+ connects 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, I'm running a server on FreeBSD-STABLE, about a week since the last make world. I was doing some stress testing with the icecast server (icecast.org) which is an open replacement for Nullsoft's shoutcast server. It comes with a stresstest app which opens a bunch of connections to the server. I found that when I opened 40 connections to the server (localhost), the machine crashed. The icecast server is running as a normal user. I haven't been able to get to the console yet so I don't know if there's more info there. I'll be investigating that shortly. At first I thought, okay, maybe a bug in the loopback driver. So I tried it from another machine on the network. Same result. Just to make sure it wasn't something specific to the icecast server, I enabled chargen in /etc/inetd.conf and did another stress test over the network. It survived 40 connections, so I tried 100. It was fine. I tried 200 and the server crashed. It seems like the server crashed when it receives many connections that require a high amount of data be sent back (icecast, which is mp3 streams or chargen). Anyone have any idea what's going on? I really hate to think that my server can be crashed this easily. I did similar stress testing with the icecast server on my laptop, running Linux 2.0.36 and it accepted 300 connections (vs. less than 40) without a problem. If Linux can do it, I know FreeBSD can. Hope someone has some ideas. Thanks in advance. --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13: 7:58 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 0BADD1557D; Thu, 25 Mar 1999 13:07:50 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id TAA13401; Thu, 25 Mar 1999 19:53:29 +0100 From: Luigi Rizzo Message-Id: <199903251853.TAA13401@labinfo.iet.unipi.it> Subject: Re: 3.1-STABLE dies on 40+ connects To: j@lumiere.net (Jesse) Date: Thu, 25 Mar 1999 19:53:29 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG In-Reply-To: from "Jesse" at Mar 25, 99 01:02:14 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1818 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Hi, > > I'm running a server on FreeBSD-STABLE, about a week since the last make > world. > > I was doing some stress testing with the icecast server (icecast.org) aren't you running out of mbufs by chance ? (it should not crash, but increasing the value could help...) luigi > which is an open replacement for Nullsoft's shoutcast server. It comes > with a stresstest app which opens a bunch of connections to the server. > > I found that when I opened 40 connections to the server (localhost), the > machine crashed. The icecast server is running as a normal user. I haven't > been able to get to the console yet so I don't know if there's more info > there. I'll be investigating that shortly. > > At first I thought, okay, maybe a bug in the loopback driver. So I tried > it from another machine on the network. Same result. > > Just to make sure it wasn't something specific to the icecast server, I > enabled chargen in /etc/inetd.conf and did another stress test over the > network. It survived 40 connections, so I tried 100. It was fine. I tried > 200 and the server crashed. > > It seems like the server crashed when it receives many connections that > require a high amount of data be sent back (icecast, which is mp3 streams > or chargen). > > Anyone have any idea what's going on? I really hate to think that my > server can be crashed this easily. I did similar stress testing with the > icecast server on my laptop, running Linux 2.0.36 and it accepted 300 > connections (vs. less than 40) without a problem. If Linux can do it, I > know FreeBSD can. > > Hope someone has some ideas. Thanks in advance. > > --- > Jesse > http://www.lumiere.net/ > > > > 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 Mar 25 13:10: 7 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id A84391557D; Thu, 25 Mar 1999 13:10:05 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id NAA00693; Thu, 25 Mar 1999 13:09:41 -0800 (PST) Date: Thu, 25 Mar 1999 13:09:41 -0800 (PST) From: Jesse To: Luigi Rizzo Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects In-Reply-To: <199903251853.TAA13401@labinfo.iet.unipi.it> 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 > aren't you running out of mbufs by chance ? (it should not crash, but > increasing the value could help...) It definitely does crash. Any recommended value for mbufs and/or any other settings while I'm recompiling the kernel? --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13:14:18 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 6201B153F6; Thu, 25 Mar 1999 13:14:12 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id TAA13450; Thu, 25 Mar 1999 19:59:51 +0100 From: Luigi Rizzo Message-Id: <199903251859.TAA13450@labinfo.iet.unipi.it> Subject: Re: 3.1-STABLE dies on 40+ connects To: j@lumiere.net (Jesse) Date: Thu, 25 Mar 1999 19:59:50 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG In-Reply-To: from "Jesse" at Mar 25, 99 01:09:22 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 292 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > aren't you running out of mbufs by chance ? (it should not crash, but > > increasing the value could help...) > > It definitely does crash. Any recommended value for mbufs and/or any other > settings while I'm recompiling the kernel? i generally use options "NMBCLUSTERS=4096" luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13:17:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id B379A14D1D; Thu, 25 Mar 1999 13:17:39 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id UAA13471; Thu, 25 Mar 1999 20:03:16 +0100 From: Luigi Rizzo Message-Id: <199903251903.UAA13471@labinfo.iet.unipi.it> Subject: Re: 3.1-STABLE dies on 40+ connects To: j@lumiere.net (Jesse) Date: Thu, 25 Mar 1999 20:03:16 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG In-Reply-To: from "Jesse" at Mar 25, 99 01:09:22 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 471 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > aren't you running out of mbufs by chance ? (it should not crash, but > > increasing the value could help...) > > It definitely does crash. i trust you -- i suspect there is some piece of the code somewhere which does not check for mcopy/mpullup/etc failures. in fact it would be nice to know if there is some reproducible way to trigger these crashes because i think this is a problem that ought to be fixed in a better way than overallocating resources. luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13:19:16 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 6A96714EC8; Thu, 25 Mar 1999 13:19:13 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id NAA00906; Thu, 25 Mar 1999 13:18:51 -0800 (PST) Date: Thu, 25 Mar 1999 13:18:51 -0800 (PST) From: Jesse To: Luigi Rizzo Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects In-Reply-To: <199903251903.UAA13471@labinfo.iet.unipi.it> 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 > > It definitely does crash. > > i trust you -- i suspect there is some piece of the code somewhere > which does not check for mcopy/mpullup/etc failures. Hmm. > in fact it would be nice to know if there is some reproducible way to > trigger these crashes because i think this is a problem that ought to > be fixed in a better way than overallocating resources. I agree. I can definitely reproduce the crash at will (at least, with current kernel settings, I'm recompiling now). All I have to do is start the icecast server and open 40 connections, or enable chargen and open 200 connections. --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 13:32:19 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 1065414D03; Thu, 25 Mar 1999 13:32:17 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id NAA04145; Thu, 25 Mar 1999 13:31:53 -0800 (PST) Date: Thu, 25 Mar 1999 13:31:53 -0800 (PST) From: Jesse To: Luigi Rizzo Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 Just to provide a little bit more info, Alex, one of the developers of icecast said that a person running OpenBSD had kernel panics under similar conditions but that it was never resolved. The machine is remote, but I can have someone stand by and record the kernel panic message while I crash it, if you it'd be useful (probably). --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14: 3:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 358AE14D21; Thu, 25 Mar 1999 14:03:35 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id OAA00423; Thu, 25 Mar 1999 14:03:08 -0800 (PST) Date: Thu, 25 Mar 1999 14:03:08 -0800 (PST) From: Jesse To: Luigi Rizzo Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects In-Reply-To: <199903251903.UAA13471@labinfo.iet.unipi.it> 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 trust you -- i suspect there is some piece of the code somewhere > which does not check for mcopy/mpullup/etc failures. > > in fact it would be nice to know if there is some reproducible way to > trigger these crashes because i think this is a problem that ought to > be fixed in a better way than overallocating resources. Okay, I updated the NMBCLUSTERS to 4096. This allowed me to get 40 clients on successfully. Here's an netstat -m with 40 clients connected: leaf:~# netstat -m 1076/1792 mbufs in use: 1025 mbufs allocated to data 51 mbufs allocated to packet headers 1023/1620/4096 mbuf clusters in use (current/peak/max) 3464 Kbytes allocated to network (62% in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines So I tried 60 clients. It crashed after the 56 client connected. I was doing netstat -m the entire time until the moment it crashed. The last one showed: leaf:~# netstat -m 4637/4704 mbufs in use: 4565 mbufs allocated to data 72 mbufs allocated to packet headers 4564/4604/4096 mbuf clusters in use (current/peak/max) 9796 Kbytes allocated to network (99% in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines So it definitely appears to be an mbuf issue. Is it normal for something to use mbuf's so quickly? Each client is being sent a 128kbps stream. I know sites like ftp.cdrom.com transfer MUCH more than this per second.. soo.. It'd of course be nicer to see FreeBSD be 'fixed' to not crash in this situation, but is there anything that the authors of icecast can do to help reduce their mbuf usage? Lastly, is it safe to raise my nmbclusters higher than 4096? Like 10000? What are the downsides? --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14: 8:25 1999 Delivered-To: freebsd-net@freebsd.org Received: from hp9000.chc-chimes.com (hp9000.chc-chimes.com [206.67.97.84]) by hub.freebsd.org (Postfix) with ESMTP id 2293E14E37; Thu, 25 Mar 1999 14:08:20 -0800 (PST) (envelope-from billf@chc-chimes.com) Received: from localhost by hp9000.chc-chimes.com with SMTP (1.39.111.2/16.2) id AA121217578; Thu, 25 Mar 1999 16:32:58 -0500 Date: Thu, 25 Mar 1999 16:32:58 -0500 (EST) From: Bill Fumerola To: Jesse Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 [ this really shouldn't be x-posted ] On Thu, 25 Mar 1999, Jesse wrote: > > i trust you -- i suspect there is some piece of the code somewhere > > which does not check for mcopy/mpullup/etc failures. > > > > in fact it would be nice to know if there is some reproducible way to > > trigger these crashes because i think this is a problem that ought to > > be fixed in a better way than overallocating resources. > > Okay, I updated the NMBCLUSTERS to 4096. This allowed me to get 40 clients > on successfully. Here's an netstat -m with 40 clients connected: > > leaf:~# netstat -m > 1076/1792 mbufs in use: > 1025 mbufs allocated to data > 51 mbufs allocated to packet headers > 1023/1620/4096 mbuf clusters in use (current/peak/max) > 3464 Kbytes allocated to network (62% in use) Wow that's high. > > So I tried 60 clients. It crashed after the 56 client connected. I was > doing netstat -m the entire time until the moment it crashed. The last one > showed: > > leaf:~# netstat -m > 4637/4704 mbufs in use: > 4565 mbufs allocated to data > 72 mbufs allocated to packet headers > 4564/4604/4096 mbuf clusters in use (current/peak/max) > 9796 Kbytes allocated to network (99% in use) Crunch. > So it definitely appears to be an mbuf issue. Definitely. Could you get some statistics on mbuf/client, for instance, which we could use to determine if the climb is linear or exponential. > Is it normal for something to use mbuf's so quickly? Each client is being > sent a 128kbps stream. I know sites like ftp.cdrom.com transfer MUCH more > than this per second.. soo.. This is certainly a bug of some sort. > It'd of course be nicer to see FreeBSD be 'fixed' to not crash in this > situation, but is there anything that the authors of icecast can do to > help reduce their mbuf usage? Does the machine panic or just crash, that would at least be the first step is just to get FreeBSD to panic. > Lastly, is it safe to raise my nmbclusters higher than 4096? Like 10000? > What are the downsides? Death or memory problems I can't remember which. - 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 Mar 25 14:10:42 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 D9AD514D9D; Thu, 25 Mar 1999 14:10:40 -0800 (PST) (envelope-from julian@whistle.com) Received: from current1.whistle.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with SMTP id OAA10737; Thu, 25 Mar 1999 14:09:58 -0800 (PST) Date: Thu, 25 Mar 1999 14:09:57 -0800 (PST) From: Julian Elischer To: Jesse Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 Thu, 25 Mar 1999, Jesse wrote: > > Just to provide a little bit more info, Alex, one of the developers of > icecast said that a person running OpenBSD had kernel panics under similar > conditions but that it was never resolved. > > The machine is remote, but I can have someone stand by and record the > kernel panic message while I crash it, if you it'd be useful (probably). the corefile would be better.... :-) > > --- > Jesse > http://www.lumiere.net/ > > > > 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 Mar 25 14:11:50 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id E0117152F3; Thu, 25 Mar 1999 14:11:48 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id OAA00583; Thu, 25 Mar 1999 14:11:29 -0800 (PST) Date: Thu, 25 Mar 1999 14:11:29 -0800 (PST) From: Jesse To: Julian Elischer Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 > > The machine is remote, but I can have someone stand by and record the > > kernel panic message while I crash it, if you it'd be useful (probably). > > the corefile would be better.... :-) How would I go about getting this? Do I have to specify some special coredump location or something? --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:20:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [209.157.86.2]) by hub.freebsd.org (Postfix) with ESMTP id C801314E06; Thu, 25 Mar 1999 14:20:36 -0800 (PST) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.9.3/8.9.1) id OAA03456; Thu, 25 Mar 1999 14:20:13 -0800 (PST) (envelope-from dillon) Date: Thu, 25 Mar 1999 14:20:13 -0800 (PST) From: Matthew Dillon Message-Id: <199903252220.OAA03456@apollo.backplane.com> To: Bill Fumerola Cc: Jesse , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects References: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org netstat -tn output would be useful -- figure out what is eating so much memory. You should be able to raise NMBCLUSTERS to 8192 easily, and possibly even higher. -Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:27:42 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 BC7AB1511D; Thu, 25 Mar 1999 14:10:44 -0800 (PST) (envelope-from julian@whistle.com) Received: from current1.whistle.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with SMTP id OAA10713; Thu, 25 Mar 1999 14:08:56 -0800 (PST) Date: Thu, 25 Mar 1999 14:08:55 -0800 (PST) From: Julian Elischer To: Jesse Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 can you get a core-dump? (and a kernel compiled with config -g to supply symbols) On Thu, 25 Mar 1999, Jesse wrote: > > > > It definitely does crash. > > > > i trust you -- i suspect there is some piece of the code somewhere > > which does not check for mcopy/mpullup/etc failures. > > Hmm. > > > in fact it would be nice to know if there is some reproducible way to > > trigger these crashes because i think this is a problem that ought to > > be fixed in a better way than overallocating resources. > > I agree. I can definitely reproduce the crash at will (at least, with > current kernel settings, I'm recompiling now). All I have to do is start > the icecast server and open 40 connections, or enable chargen and open 200 > connections. > > --- > Jesse > http://www.lumiere.net/ > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-hackers" 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 Mar 25 14:29:53 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 8268715109; Thu, 25 Mar 1999 14:29:51 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id OAA01169; Thu, 25 Mar 1999 14:29:28 -0800 (PST) Date: Thu, 25 Mar 1999 14:29:28 -0800 (PST) From: Jesse To: Bill Fumerola Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 > Definitely. Could you get some statistics on mbuf/client, for instance, > which we could use to determine if the climb is linear or exponential. Okay, I got my stats. Keep in mind that in each case, the mbuf's would vary +-200 but I tried to get an average value. Also, particularly in icecast's case, the number of mbuf's used would be higher in the first 20 seconds of connection, then drop down some. I also compared shoutcast, another program which does the same thing. It spawns a child for each client through (as opposed to a single process). ICECAST -- 10 Clients 281/1752/4096 mbuf clusters in use (current/peak/max) 3760 Kbytes allocated to network (16% in use) ICECAST -- 20 Clients 513/1752/4096 mbuf clusters in use (current/peak/max) 3760 Kbytes allocated to network (29% in use) ICECAST -- 30 Clients 860/1752/4096 mbuf clusters in use (current/peak/max) 3760 Kbytes allocated to network (48% in use) ICECAST -- 40 Clients 1023/1620/4096 mbuf clusters in use (current/peak/max) 3464 Kbytes allocated to network (62% in use) ICECSAT -- 50 Clients 3340/3884/4096 mbuf clusters in use (current/peak/max) 8268 Kbytes allocated to network (85% in use) ICECAST -- 60 Clients (crash at 56) 4564/4604/4096 mbuf clusters in use (current/peak/max) 9796 Kbytes allocated to network (99% in use) SHOUTCAST -- 40 Clients 435/702/4096 mbuf clusters in use (current/peak/max) 1528 Kbytes allocated to network (61% in use) SHOUTCAST -- 60 Clients 843/980/4096 mbuf clusters in use (current/peak/max) 2112 Kbytes allocated to network (86% in use) SHOUTCAST -- 100 Clients 1717/1752/4096 mbuf clusters in use (current/peak/max) 3760 Kbytes allocated to network (97% in use) > > Is it normal for something to use mbuf's so quickly? Each client is being > > sent a 128kbps stream. I know sites like ftp.cdrom.com transfer MUCH more > > than this per second.. soo.. > > This is certainly a bug of some sort. Seeing them compared to shoutcast's, I'd tend to agree. Still, crashing really sucks. :/ > Does the machine panic or just crash, that would at least be the first > step is just to get FreeBSD to panic. I don't know, the machine is remote. I think I'm going to figure out how to get a panic core dump, then have someone watch it and record the info when it dies. --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:32: 5 1999 Delivered-To: freebsd-net@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [209.157.86.2]) by hub.freebsd.org (Postfix) with ESMTP id D63EE14EA8; Thu, 25 Mar 1999 14:31:57 -0800 (PST) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.9.3/8.9.1) id OAA03635; Thu, 25 Mar 1999 14:31:28 -0800 (PST) (envelope-from dillon) Date: Thu, 25 Mar 1999 14:31:28 -0800 (PST) From: Matthew Dillon Message-Id: <199903252231.OAA03635@apollo.backplane.com> To: Jesse Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects References: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org :Just to provide a little bit more info, Alex, one of the developers of :icecast said that a person running OpenBSD had kernel panics under similar :conditions but that it was never resolved. : :The machine is remote, but I can have someone stand by and record the :kernel panic message while I crash it, if you it'd be useful (probably). : :--- :Jesse :http://www.lumiere.net/ On the face of it it sounds to me like icecast is perhaps making the socket buffers bigger --- maybe too big. The default socket buffer size is 16K receiving, 16K sending. 40 x 16K x 2 = 1MB. If icecast is increasing the size of the socket buffers, this value can explode. netstat -tn on a running system ought to give you some idea about that by seeing how backed-up the socket buffers get. -Matt Matthew Dillon To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:41:16 1999 Delivered-To: freebsd-net@freebsd.org Received: from hp9000.chc-chimes.com (hp9000.chc-chimes.com [206.67.97.84]) by hub.freebsd.org (Postfix) with ESMTP id D328814A23; Thu, 25 Mar 1999 14:41:09 -0800 (PST) (envelope-from billf@chc-chimes.com) Received: from localhost by hp9000.chc-chimes.com with SMTP (1.39.111.2/16.2) id AA133919608; Thu, 25 Mar 1999 17:06:48 -0500 Date: Thu, 25 Mar 1999 17:06:48 -0500 (EST) From: Bill Fumerola To: Jesse Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 Thu, 25 Mar 1999, Jesse wrote: > I also compared shoutcast, another program which does the same thing. It > spawns a child for each client through (as opposed to a single process). [ some number crunching on my TI-85 ...] If you think that the growth is linear (I don't) y = 85x - 1242 If you think the growth is exponential (I do, this fits a lot nicer.) y = 2.2x^2 - 75x + 900 Where x is the amount of clients. and y is the mbufs in use. I don't know if that helps any developers, but it sure was fun. - 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 Mar 25 14:42:40 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id EC38314EF5; Thu, 25 Mar 1999 14:42:38 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id OAA01393; Thu, 25 Mar 1999 14:42:11 -0800 (PST) Date: Thu, 25 Mar 1999 14:42:11 -0800 (PST) From: Jesse To: Matthew Dillon Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects In-Reply-To: <199903252231.OAA03635@apollo.backplane.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, 25 Mar 1999, Matthew Dillon wrote: > On the face of it it sounds to me like icecast is perhaps making the > socket buffers bigger --- maybe too big. The default socket buffer > size is 16K receiving, 16K sending. 40 x 16K x 2 = 1MB. If > icecast is increasing the size of the socket buffers, this value can > explode. netstat -tn on a running system ought to give you some idea > about that by seeing how backed-up the socket buffers get. That certainly seems like the case. Assuming this is the culprit, is it still worthwhile to get the coredump/panic info to see why it's crashing? Here's the output of netstat -tn with 43 clients connected (icecast): leaf:~# netstat -tn Active Internet connections Proto Recv-Q Send-Q Local Address Foreign Address (state) tcp 0 28086 207.218.152.15.8000 206.170.14.10.2229 ESTABLISHED tcp 0 28086 207.218.152.15.8000 206.170.14.10.2228 ESTABLISHED tcp 0 28086 207.218.152.15.8000 206.170.14.10.2227 ESTABLISHED tcp 0 28086 207.218.152.15.8000 206.170.14.10.2226 ESTABLISHED tcp 0 41181 207.218.152.15.8000 206.170.14.10.2225 ESTABLISHED tcp 0 41181 207.218.152.15.8000 206.170.14.10.2219 ESTABLISHED tcp 0 47021 207.218.152.15.8000 206.170.14.10.2210 ESTABLISHED tcp 0 29456 207.218.152.15.8000 206.170.14.10.2209 ESTABLISHED tcp 0 50830 207.218.152.15.8000 206.170.14.10.2208 ESTABLISHED tcp 0 98053 207.218.152.15.8000 206.26.192.14.57940 ESTABLISHED tcp 0 53867 207.218.152.15.8000 128.111.82.59.1516 ESTABLISHED tcp 0 61915 207.218.152.15.8000 128.151.43.91.1799 ESTABLISHED <... lots more similar output chopped ...> And for comparison, shoutcast with 40 clients: leaf:~# netstat -tn Active Internet connections Proto Recv-Q Send-Q Local Address Foreign Address (state) tcp 0 0 207.218.152.15.8000 206.170.14.10.2280 ESTABLISHED tcp 0 7646 207.218.152.15.8000 206.170.14.10.2279 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2278 ESTABLISHED tcp 0 0 207.218.152.15.8000 206.170.14.10.2277 ESTABLISHED tcp 0 4726 207.218.152.15.8000 206.170.14.10.2276 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2275 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2274 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2273 ESTABLISHED tcp 0 6186 207.218.152.15.8000 206.170.14.10.2272 ESTABLISHED tcp 0 0 207.218.152.15.8000 206.170.14.10.2271 ESTABLISHED tcp 0 2875 207.218.152.15.8000 206.170.14.10.2270 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2269 ESTABLISHED tcp 0 16015 207.218.152.15.8000 206.170.14.10.2268 ESTABLISHED <... yet more chopped ...> --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:46:50 1999 Delivered-To: freebsd-net@freebsd.org Received: from biggusdiskus.flyingfox.com (parker-T1-2-gw.sf3d.best.net [209.157.165.30]) by hub.freebsd.org (Postfix) with ESMTP id BB53314EF5 for ; Thu, 25 Mar 1999 14:46:49 -0800 (PST) (envelope-from jas@flyingfox.com) Received: (from jas@localhost) by biggusdiskus.flyingfox.com (8.8.8/8.8.5) id PAA21856; Thu, 25 Mar 1999 15:57:54 -0800 (PST) Date: Thu, 25 Mar 1999 15:57:54 -0800 (PST) From: Jim Shankland Message-Id: <199903252357.PAA21856@biggusdiskus.flyingfox.com> To: dillon@apollo.backplane.com, j@lumiere.net Subject: Re: 3.1-STABLE dies on 40+ connects Cc: freebsd-net@FreeBSD.ORG In-Reply-To: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Bingo. Look at those send buffers. Of course, crashing when you run out of mbuf clusters is not what I'd call graceful degradation. Jim Shankland NLynx Systems, Inc. [Redirected to -net only; I'm getting tired of seeing each of these messages twice.] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 14:47:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from hp9000.chc-chimes.com (hp9000.chc-chimes.com [206.67.97.84]) by hub.freebsd.org (Postfix) with ESMTP id 3087114FC5; Thu, 25 Mar 1999 14:47:22 -0800 (PST) (envelope-from billf@chc-chimes.com) Received: from localhost by hp9000.chc-chimes.com with SMTP (1.39.111.2/16.2) id AA135779963; Thu, 25 Mar 1999 17:12:43 -0500 Date: Thu, 25 Mar 1999 17:12:43 -0500 (EST) From: Bill Fumerola To: Jesse Cc: Matthew Dillon , Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 Thu, 25 Mar 1999, Jesse wrote: > leaf:~# netstat -tn > Active Internet connections > Proto Recv-Q Send-Q Local Address Foreign Address (state) > tcp 0 28086 207.218.152.15.8000 206.170.14.10.2229 > ESTABLISHED > tcp 0 28086 207.218.152.15.8000 206.170.14.10.2228 > ESTABLISHED > tcp 0 28086 207.218.152.15.8000 206.170.14.10.2227 > ESTABLISHED > tcp 0 28086 207.218.152.15.8000 206.170.14.10.2226 > ESTABLISHED > tcp 0 41181 207.218.152.15.8000 206.170.14.10.2225 > ESTABLISHED > tcp 0 41181 207.218.152.15.8000 206.170.14.10.2219 > ESTABLISHED > tcp 0 47021 207.218.152.15.8000 206.170.14.10.2210 > ESTABLISHED > tcp 0 29456 207.218.152.15.8000 206.170.14.10.2209 > ESTABLISHED > tcp 0 50830 207.218.152.15.8000 206.170.14.10.2208 > ESTABLISHED > tcp 0 98053 207.218.152.15.8000 206.26.192.14.57940 ^^^^^ > ESTABLISHED > tcp 0 53867 207.218.152.15.8000 128.111.82.59.1516 > ESTABLISHED > tcp 0 61915 207.218.152.15.8000 128.151.43.91.1799 ^^^^^ All of these are _really_ backed up. > Active Internet connections > Proto Recv-Q Send-Q Local Address Foreign Address (state) > tcp 0 0 207.218.152.15.8000 206.170.14.10.2280 > ESTABLISHED > tcp 0 7646 207.218.152.15.8000 206.170.14.10.2279 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2278 > ESTABLISHED > tcp 0 0 207.218.152.15.8000 206.170.14.10.2277 > ESTABLISHED > tcp 0 4726 207.218.152.15.8000 206.170.14.10.2276 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2275 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2274 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2273 > ESTABLISHED > tcp 0 6186 207.218.152.15.8000 206.170.14.10.2272 > ESTABLISHED > tcp 0 0 207.218.152.15.8000 206.170.14.10.2271 > ESTABLISHED > tcp 0 2875 207.218.152.15.8000 206.170.14.10.2270 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2269 > ESTABLISHED > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2268 > ESTABLISHED These seem to be handling the data in a timely fashion. - 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 Mar 25 14:54:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 0D29D14C2F; Thu, 25 Mar 1999 14:54:34 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id OAA01491; Thu, 25 Mar 1999 14:54:08 -0800 (PST) Date: Thu, 25 Mar 1999 14:54:08 -0800 (PST) From: Jesse To: Bill Fumerola Cc: Matthew Dillon , Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 > > tcp 0 98053 207.218.152.15.8000 206.26.192.14.57940 > ^^^^^ > > ESTABLISHED > > tcp 0 61915 207.218.152.15.8000 128.151.43.91.1799 > ^^^^^ > All of these are _really_ backed up. Well, this is over an ethernet connection. 40 clients, each doing 128kbps clogs it pretty bad. =) But it confirms Matt Dillon's theory that the buffers have been made larger than 16k. > > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2268 > > ESTABLISHED > > These seem to be handling the data in a timely fashion. Same overloaded ethernet link. But they're all <= 16k. --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 15: 6:33 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 1CF7914EF5; Thu, 25 Mar 1999 15:06:29 -0800 (PST) (envelope-from julian@whistle.com) Received: from current1.whistle.com (current1.whistle.com [207.76.205.22]) by alpo.whistle.com (8.9.1a/8.9.1) with SMTP id PAA13093; Thu, 25 Mar 1999 15:04:59 -0800 (PST) Date: Thu, 25 Mar 1999 15:04:58 -0800 (PST) From: Julian Elischer To: Bill Fumerola Cc: Jesse , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects 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 The generation of traffic is linear with #clinets and the drain is a constant, (in fact it may even decrease with contention) julian On Thu, 25 Mar 1999, Bill Fumerola wrote: > On Thu, 25 Mar 1999, Jesse wrote: > > > I also compared shoutcast, another program which does the same thing. It > > spawns a child for each client through (as opposed to a single process). > > [ some number crunching on my TI-85 ...] > > If you think that the growth is linear (I don't) > y = 85x - 1242 > > If you think the growth is exponential (I do, this fits a lot nicer.) > y = 2.2x^2 - 75x + 900 > > Where x is the amount of clients. and y is the mbufs in use. > I don't know if that helps any developers, but it sure was fun. > > - 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-hackers" 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 Mar 25 15: 7:10 1999 Delivered-To: freebsd-net@freebsd.org Received: from biggusdiskus.flyingfox.com (parker-T1-2-gw.sf3d.best.net [209.157.165.30]) by hub.freebsd.org (Postfix) with ESMTP id 61FCB150AA for ; Thu, 25 Mar 1999 15:07:08 -0800 (PST) (envelope-from jas@flyingfox.com) Received: (from jas@localhost) by biggusdiskus.flyingfox.com (8.8.8/8.8.5) id QAA21926; Thu, 25 Mar 1999 16:18:12 -0800 (PST) Date: Thu, 25 Mar 1999 16:18:12 -0800 (PST) From: Jim Shankland Message-Id: <199903260018.QAA21926@biggusdiskus.flyingfox.com> To: billf@chc-chimes.com, j@lumiere.net Subject: Re: 3.1-STABLE dies on 40+ connects Cc: freebsd-net@FreeBSD.ORG In-Reply-To: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > All of these are _really_ backed up. > > Well, this is over an ethernet connection. 40 clients, each doing 128kbps > clogs it pretty bad. =) But it confirms Matt Dillon's theory that the > buffers have been made larger than 16k. It's not the Ethernet that's backed up. The bytes stay in the send buffers until they're acked by the receiving end. It's trivial for a program to fill the send buffer to its limit; you just have to be able to do write() calls faster than data can be delivered and ACKed by the remote end. > > > tcp 0 16015 207.218.152.15.8000 206.170.14.10.2268 > > > ESTABLISHED > > > > These seem to be handling the data in a timely fashion. > > Same overloaded ethernet link. But they're all <= 16k. The first program has explicitly increased the size of its send buffers with setsockopt(). The second has not, and gets the default size. In both cases, the send buffers are (near) full. The real problem is that the write() calls should be stalling or returning ENOBUFS (preferably the former) when no more mbuf clusters are available. Jim Shankland NLynx Systems, Inc. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 15:25:52 1999 Delivered-To: freebsd-net@freebsd.org Received: from noc.demon.net (server.noc.demon.net [193.195.224.4]) by hub.freebsd.org (Postfix) with ESMTP id A74BC15265 for ; Thu, 25 Mar 1999 15:25:39 -0800 (PST) (envelope-from fanf@demon.net) Received: by noc.demon.net; id XAA04603; Thu, 25 Mar 1999 23:25:18 GMT Received: from fanf.noc.demon.net(195.11.55.83) by inside.noc.demon.net via smap (3.2) id xma004570; Thu, 25 Mar 99 23:25:09 GMT Received: from fanf by fanf.noc.demon.net with local (Exim 1.73 #2) id 10QJUr-0003NB-00; Thu, 25 Mar 1999 23:25:09 +0000 To: net@freebsd.org From: Tony Finch Subject: Re: IP alias configuration for high availability Newsgroups: chiark.mail.freebsd.net In-Reply-To: <4.1.19990323134009.0099fe10@mail.dnai.com> Organization: Deliberate Obfuscation To Amuse Tony Message-Id: Date: Thu, 25 Mar 1999 23:25:09 +0000 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Mike Thompson wrote: > >The idea is that first network address for a server would be >fixed on the reserved IP addresses of 192.168.xxx.xxx. These >fixed addresses would be used for private communication >between the various servers. > >The other addresses used by each server would be from the >Internet IP address block assigned from our ISP (example >209.185.152.1 thru 209.185.152.20). > >With such a configuration we hope to be able to hot swap >servers among the various Internet IP addresses for system >maintenance and emergency situations for high availability. > >From my testing it seems everything works pretty well. I have >found that I need to manually break the association between >an IP address and ethernet address on the servers by using the >"arp" utility. You can avoid the need to fiddle with arp and ifconfig by setting up the alias on the server's loopback interface. You then control which servers are live by changing the routes on the router; the servers don't need to be told what's going on. To set up a server: ifconfig ed0 inet 192.168.0.1 netmask 255.255.255.0 ifconfig lo0 inet 209.185.152.1 netmask 255.255.255.255 alias On the router: route add -host 209.185.152.1 192.168.0.1 Tony. -- f.a.n.finch dot@dotat.at fanf@demon.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 15:26:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id A792715554; Thu, 25 Mar 1999 15:26:25 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id PAA02048; Thu, 25 Mar 1999 15:26:04 -0800 (PST) Date: Thu, 25 Mar 1999 15:26:04 -0800 (PST) From: Jesse To: Matthew Dillon Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects (resolved) In-Reply-To: <199903252231.OAA03635@apollo.backplane.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 the face of it it sounds to me like icecast is perhaps making the > socket buffers bigger --- maybe too big. The default socket buffer I talked to the icecast developers. A simple #undef SOCKET_OPERATIONS 1 stops it from changing the buffer size and they'll be incorporating that change (for *BSD systems). New testing with #undef SOCKET_OPERATIONS 1: I have 121 clients connected and mbuf usage hovers bewteen 500-1100 (as opposed to where it'd die at 4096 with 56 before). Additionally, if I cut it down to 61 clients the mbuf is in the 50-600 range (usually sitting around 90-150 range). I suspect the exponential increase at 91 and 121 clients is because of the massive amount of bandwidth (even for lo0), i.e., around 22Mbps. Thanks everybody for your help! FreeBSD is great and it's 80% about the community. Not that I didn't always know it. =) --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 15:43:10 1999 Delivered-To: freebsd-net@freebsd.org Received: from biggusdiskus.flyingfox.com (parker-T1-2-gw.sf3d.best.net [209.157.165.30]) by hub.freebsd.org (Postfix) with ESMTP id ADAEB14D64 for ; Thu, 25 Mar 1999 15:43:09 -0800 (PST) (envelope-from jas@flyingfox.com) Received: (from jas@localhost) by biggusdiskus.flyingfox.com (8.8.8/8.8.5) id QAA22060; Thu, 25 Mar 1999 16:54:16 -0800 (PST) Date: Thu, 25 Mar 1999 16:54:16 -0800 (PST) From: Jim Shankland Message-Id: <199903260054.QAA22060@biggusdiskus.flyingfox.com> To: j@lumiere.net Subject: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org A thought related to this discussion: does it make sense to allow the send buffers to be larger than the peer's advertised window size? In other words, why "preposition" those bytes in the kernel before the peer has indicated a willingness to accept them? Jim Shankland NLynx Systems, Inc. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 16: 5:20 1999 Delivered-To: freebsd-net@freebsd.org Received: from peach.ocn.ne.jp (peach.ocn.ne.jp [210.145.254.87]) by hub.freebsd.org (Postfix) with ESMTP id 982CF14C47; Thu, 25 Mar 1999 16:05:14 -0800 (PST) (envelope-from dcs@newsguy.com) Received: from newsguy.com by peach.ocn.ne.jp (8.9.1a/OCN) id JAA03966; Fri, 26 Mar 1999 09:05:02 +0900 (JST) Message-ID: <36FACE5B.D0926C2C@newsguy.com> Date: Fri, 26 Mar 1999 09:01:31 +0900 From: "Daniel C. Sobral" X-Mailer: Mozilla 4.5 [en] (Win98; I) X-Accept-Language: pt-BR,ja MIME-Version: 1.0 To: Jesse Cc: Matthew Dillon , Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects (resolved) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Jesse wrote: > > Thanks everybody for your help! FreeBSD is great and it's 80% about the > community. Not that I didn't always know it. =) Notice that a core dump + kernel with symbols would still be very useful to find *where* in our code is the bug. -- Daniel C. Sobral (8-DCS) dcs@newsguy.com dcs@freebsd.org "What kind of psychologist laughs at her patients?" "I don't laugh at all of them." To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 16:10:11 1999 Delivered-To: freebsd-net@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id 3C3AB1530F; Thu, 25 Mar 1999 16:10:09 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id QAA02512; Thu, 25 Mar 1999 16:09:44 -0800 (PST) Date: Thu, 25 Mar 1999 16:09:44 -0800 (PST) From: Jesse To: "Daniel C. Sobral" Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects (resolved) In-Reply-To: <36FACE5B.D0926C2C@newsguy.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 > > Thanks everybody for your help! FreeBSD is great and it's 80% about the > > community. Not that I didn't always know it. =) > > Notice that a core dump + kernel with symbols would still be very > useful to find *where* in our code is the bug. I asked in other messages if anyone still wanted me to do that. No one replied. Anyway, I'll do this at an off hour and get a core dump. --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 16:51:48 1999 Delivered-To: freebsd-net@freebsd.org Received: from gecko.nas.nasa.gov (gecko.nas.nasa.gov [129.99.50.17]) by hub.freebsd.org (Postfix) with ESMTP id BBBBE14E95 for ; Thu, 25 Mar 1999 16:51:43 -0800 (PST) (envelope-from kml@gecko.nas.nasa.gov) Received: from gecko.nas.nasa.gov (kml@localhost) by gecko.nas.nasa.gov (8.9.1a/NAS8.8.7n) with ESMTP id QAA17517; Thu, 25 Mar 1999 16:51:09 -0800 (PST) Message-Id: <199903260051.QAA17517@gecko.nas.nasa.gov> To: Jim Shankland Cc: j@lumiere.net, freebsd-net@FreeBSD.ORG Subject: Re: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) In-reply-to: Your message of "Thu, 25 Mar 1999 16:54:16 PST." <199903260054.QAA22060@biggusdiskus.flyingfox.com> Date: Thu, 25 Mar 1999 16:51:09 -0800 From: "Kevin M. Lahey" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In message <199903260054.QAA22060@biggusdiskus.flyingfox.com>, Jim Shankland writes: >A thought related to this discussion: does it make sense to allow the >send buffers to be larger than the peer's advertised window size? >In other words, why "preposition" those bytes in the kernel before >the peer has indicated a willingness to accept them? Interestingly enough, no memory is actually used until data arrives. The socket buffer size is merely a cap on the amount of memory that could possibly be allocated for that connection. Kevin kml@nas.nasa.gov To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 17:18:16 1999 Delivered-To: freebsd-net@freebsd.org Received: from biggusdiskus.flyingfox.com (parker-T1-2-gw.sf3d.best.net [209.157.165.30]) by hub.freebsd.org (Postfix) with ESMTP id 4D1D214ED9 for ; Thu, 25 Mar 1999 17:18:07 -0800 (PST) (envelope-from jas@flyingfox.com) Received: (from jas@localhost) by biggusdiskus.flyingfox.com (8.8.8/8.8.5) id SAA22400; Thu, 25 Mar 1999 18:29:14 -0800 (PST) Date: Thu, 25 Mar 1999 18:29:14 -0800 (PST) From: Jim Shankland Message-Id: <199903260229.SAA22400@biggusdiskus.flyingfox.com> To: kml@nas.nasa.gov Subject: Re: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) Cc: freebsd-net@freebsd.org In-Reply-To: <199903260051.QAA17517@gecko.nas.nasa.gov> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org "Kevin M. Lahey" writes: > In message <199903260054.QAA22060@biggusdiskus.flyingfox.com>, > Jim Shankland writes: > >A thought related to this discussion: does it make sense to allow the > >send buffers to be larger than the peer's advertised window size? > >In other words, why "preposition" those bytes in the kernel before > >the peer has indicated a willingness to accept them? > > Interestingly enough, no memory is actually used until data arrives. > The socket buffer size is merely a cap on the amount of memory that > could possibly be allocated for that connection. I'm not sure whether you misunderstood my point, or were just making a related observation. Just in case, let me clarify with a concrete example: A program sets its socket send buffer size to 64KB. At some instant, the kernel has 8KB of data in its send buffer, which is exactly how much the receiving end has indicated (via the window advertisement) it is willing to accept. The sending process issues another 8 KB write() call. Should the kernel accept the 8 KB of data, allocating more mbufs and returning from the write(), or should it stall the write() call until the receiving end grows the window? Making this change would not fix the crash-when-you-run-out-of-mbufs problem, but would make it a bit less frequent, and would conserve kernel memory. It's just a thought, but I thought I'd run it up the flagpole .... Jim Shankland NLynx Systems, Inc. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Thu Mar 25 17:33:53 1999 Delivered-To: freebsd-net@freebsd.org Received: from gecko.nas.nasa.gov (gecko.nas.nasa.gov [129.99.50.17]) by hub.freebsd.org (Postfix) with ESMTP id 1C310154F3 for ; Thu, 25 Mar 1999 17:33:52 -0800 (PST) (envelope-from kml@gecko.nas.nasa.gov) Received: from gecko.nas.nasa.gov (kml@localhost) by gecko.nas.nasa.gov (8.9.1a/NAS8.8.7n) with ESMTP id RAA17817; Thu, 25 Mar 1999 17:33:32 -0800 (PST) Message-Id: <199903260133.RAA17817@gecko.nas.nasa.gov> To: Jim Shankland Cc: freebsd-net@freebsd.org Subject: Re: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) In-reply-to: Your message of "Thu, 25 Mar 1999 18:29:14 PST." <199903260229.SAA22400@biggusdiskus.flyingfox.com> Date: Thu, 25 Mar 1999 17:33:32 -0800 From: "Kevin M. Lahey" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In message <199903260229.SAA22400@biggusdiskus.flyingfox.com>, Jim Shankland writes: >I'm not sure whether you misunderstood my point, or were just making >a related observation. Just in case, let me clarify with a concrete >example: Sorry, I was being stupid; I was thinking about receive buffers rather than send buffers. >A program sets its socket send buffer size to 64KB. At some instant, >the kernel has 8KB of data in its send buffer, which is exactly >how much the receiving end has indicated (via the window advertisement) >it is willing to accept. The sending process issues another 8 KB write() >call. Should the kernel accept the 8 KB of data, allocating more mbufs >and returning from the write(), or should it stall the write() call >until the receiving end grows the window? Hmmm, I guess I'd object that the receiver could increase the advertised window size at any time, so it'd be nice to have at least some amount of data sitting in the buffer, ready to go. Granted, the window size is limited by the window scale factor, so there would be some sort of reasonable upper limit on the size of the buffer. The folks at PSC have done some interesting work on autotuning buffer sizes to solve exactly this problem. Check it out: http://www.psc.edu/networking/auto.html Sorry for my confusion on the earlier point, Kevin kml@nas.nasa.gov To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 0: 0:27 1999 Delivered-To: freebsd-net@freebsd.org Received: from kamna.i.cz (kamna.i.cz [193.85.255.30]) by hub.freebsd.org (Postfix) with SMTP id 5F6611554D for ; Fri, 26 Mar 1999 00:00:25 -0800 (PST) (envelope-from mm@i.cz) Received: (qmail 16845 invoked from network); 26 Mar 1999 08:00:05 -0000 Received: from woody.i.cz (@193.85.255.60) by kamna.i.cz with SMTP; 26 Mar 1999 08:00:05 -0000 Content-Length: 226 Message-ID: X-Mailer: XFMail 1.3 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-2 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199903242123.PAA11962@free.pcs> Date: Fri, 26 Mar 1999 09:00:04 +0100 (MET) Reply-To: mm@i.cz From: Martin Machacek To: freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 24-Mar-99 Jonathan Lemon wrote: > So where's the definition for layer-four switches? (What the heck > is a layer 4 switch anyway?) Layer 4 switch is a pure marketing bullshit. Martin --- [PGP KeyID F3F409C4]] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 0: 8:19 1999 Delivered-To: freebsd-net@freebsd.org Received: from silver.sms.fi (silver.sms.fi [194.111.122.17]) by hub.freebsd.org (Postfix) with SMTP id E46B51555E for ; Fri, 26 Mar 1999 00:06:28 -0800 (PST) (envelope-from pete@sms.fi) Received: from sms.fi (localhost.sms.fi [127.0.0.1]) by silver.sms.fi (8.9.2/8.9.2) with ESMTP id KAA20832; Fri, 26 Mar 1999 10:04:38 +0200 (EET) (envelope-from pete@sms.fi) Message-ID: <36FB3F96.3303181A@sms.fi> Date: Fri, 26 Mar 1999 10:04:38 +0200 From: Petri Helenius X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 3.1-STABLE i386) X-Accept-Language: en,fi MIME-Version: 1.0 To: Jim Shankland Cc: dillon@apollo.backplane.com, j@lumiere.net, freebsd-net@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects References: <199903252357.PAA21856@biggusdiskus.flyingfox.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Jim Shankland wrote: > > Bingo. Look at those send buffers. > > Of course, crashing when you run out of mbuf clusters is not what I'd > call graceful degradation. > Agreed. I've seen this behaviour on other kind of applications also, without the "need" for the application to ask for bigger buffers. Pete > Jim Shankland > NLynx Systems, Inc. > > [Redirected to -net only; I'm getting tired of seeing each of these > messages twice.] > > 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 Fri Mar 26 0: 8:21 1999 Delivered-To: freebsd-net@freebsd.org Received: from silver.sms.fi (silver.sms.fi [194.111.122.17]) by hub.freebsd.org (Postfix) with SMTP id 016A815568 for ; Fri, 26 Mar 1999 00:07:30 -0800 (PST) (envelope-from pete@sms.fi) Received: from sms.fi (localhost.sms.fi [127.0.0.1]) by silver.sms.fi (8.9.2/8.9.2) with ESMTP id KAA20841; Fri, 26 Mar 1999 10:07:06 +0200 (EET) (envelope-from pete@sms.fi) Message-ID: <36FB402A.5785497D@sms.fi> Date: Fri, 26 Mar 1999 10:07:06 +0200 From: Petri Helenius X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 3.1-STABLE i386) X-Accept-Language: en,fi MIME-Version: 1.0 To: Jim Shankland Cc: j@lumiere.net, freebsd-net@FreeBSD.ORG Subject: Re: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) References: <199903260054.QAA22060@biggusdiskus.flyingfox.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Jim Shankland wrote: > > A thought related to this discussion: does it make sense to allow the > send buffers to be larger than the peer's advertised window size? > In other words, why "preposition" those bytes in the kernel before > the peer has indicated a willingness to accept them? > If you have high bandwidth paths which low latency, the kernel can send the window faster than your process gets scheduled again. So it's useful to have multiple windows worth of send buffers. Pete To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 3:20:23 1999 Delivered-To: freebsd-net@freebsd.org Received: from noc.demon.net (server.noc.demon.net [193.195.224.4]) by hub.freebsd.org (Postfix) with ESMTP id 5F6AE14EFC for ; Fri, 26 Mar 1999 03:20:22 -0800 (PST) (envelope-from fanf@demon.net) Received: by noc.demon.net; id LAA14429; Fri, 26 Mar 1999 11:20:00 GMT Received: from fanf.noc.demon.net(195.11.55.83) by inside.noc.demon.net via smap (3.2) id xmaa14366; Fri, 26 Mar 99 11:19:44 GMT Received: from fanf by fanf.noc.demon.net with local (Exim 1.73 #2) id 10QUeN-0003eu-00; Fri, 26 Mar 1999 11:19:43 +0000 To: net@freebsd.org From: Tony Finch Subject: Re: clustering/load balancing In-Reply-To: <51315.922076863@gjp.erols.com> References: <36F1BA88.2F1CF0FB@whistle.com> Message-Id: Date: Fri, 26 Mar 1999 11:19:43 +0000 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org "Gary Palmer" wrote: >Julian Elischer wrote: >> >> +-------[Machine B] >> | >> [internet]-----[ Machine A ]-----+-------[Machine C] >> | >> +-------[Machine D] > ^^^^^^^^^ > >Single Point Of Failure A more problematic single point of failure is the shared filesystem that's presumably hanging off the right of the diagram. NetApps are fortunately quite reliable :-) Tony. -- f.a.n.finch dot@dotat.at fanf@demon.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 6: 0: 4 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 70AF6152A5 for ; Fri, 26 Mar 1999 06:00:00 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id OAA17402 for ; Fri, 26 Mar 1999 14:59:36 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@FreeBSD.org Subject: Re: Dummynet Date: Fri, 26 Mar 1999 14:42:02 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain References: <199903251351.OAA12974@labinfo.iet.unipi.it> MIME-Version: 1.0 Message-Id: <99032615015301.00294@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Am Do , 25 Mär 1999 schrieb Luigi Rizzo: ... > > > > I transmitted one file with 56940 Bytes from PC1 to PC2 via ftp > > in 13.24 sec. (4.30 KByte/s): > > > > Here is the output from "ipfw show": > > 00010 69 75877 pipe 1 ip from 192.168.0.238 to 192.168.0.96 > > 00150 0 0 allow ip from any to any via lo0 > > 00200 0 0 deny ip from any to 127.0.0.0/8 > > 65535 12 0 allow ip from any to any > > > > Here is the output from "ipfw pipe show": > > 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops > > there must be some problem somewhere, are you sure you have the latest > bridge.c ipfw.c and ip_dummynet.c and binaries are derived > from them -- it makes no sense > that you have 10 pkts in the queue and certainly > they cannot sum up to over 300Kbytes. > (it might well be my fault... can you refresh me on your FreeBSD 3.1-RELEASE, Here are the files listed I updatet: bridge.c v 1.6 Thu Feb 25 10:48:31 1999 UTC by kato if_ed.c $Id: if_ed.c,v 1.150 1999/03/17 16:44:51 luigi Exp $ ip_dummynet.c $Id: ip_dummynet.c,v 1.8 1999/01/27 22:42:24 dillon Exp $ ip_fw.c $Id: ip_fw.c,v 1.104 1999/02/16 10:49:52 dfr Exp $ Mybe I mixed some versions? > os and the version of the above files... in doubt rm kernel bridge.o > ipfw.o ip_dummynet.o from your compile dir and rebuild one.) I made allways a "make clean" in my compile directory, after I installed the new sources. what does the fields in this record exactly mean? 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops | | | | | pipenr bandwidth delay ?? ???? Luigi: Do you read mails send to ? Bye, bye, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 6:13:33 1999 Delivered-To: freebsd-net@freebsd.org Received: from rt2.synx.com (tech.boostworks.com [194.167.81.239]) by hub.freebsd.org (Postfix) with ESMTP id AEF4B14FAC for ; Fri, 26 Mar 1999 06:13:14 -0800 (PST) (envelope-from root@synx.com) Received: from synx.com (rn.synx.com [192.1.1.241]) by rt2.synx.com (8.9.1/8.9.1) with ESMTP id PAA12462; Fri, 26 Mar 1999 15:19:55 +0100 (CET) Message-Id: <199903261419.PAA12462@rt2.synx.com> Date: Fri, 26 Mar 1999 15:11:35 +0100 (CET) From: Remy Nonnenmacher Reply-To: remy@synx.com Subject: Re: Dummynet To: mbretter@sbox.tu-graz.ac.at Cc: freebsd-net@FreeBSD.ORG In-Reply-To: <99032615015301.00294@babsi.tu-graz.ac.at> MIME-Version: 1.0 Content-Type: TEXT/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8BIT Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 26 Mar, Michael Bretterklieber wrote: > Am Do , 25 Mär 1999 schrieb Luigi Rizzo: > ... >> > >> > I transmitted one file with 56940 Bytes from PC1 to PC2 via ftp >> > in 13.24 sec. (4.30 KByte/s): >> > >> > Here is the output from "ipfw show": >> > 00010 69 75877 pipe 1 ip from 192.168.0.238 to 192.168.0.96 >> > 00150 0 0 allow ip from any to any via lo0 >> > 00200 0 0 deny ip from any to 127.0.0.0/8 >> > 65535 12 0 allow ip from any to any >> > >> > Here is the output from "ipfw pipe show": >> > 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops >> >> there must be some problem somewhere, are you sure you have the latest >> bridge.c ipfw.c and ip_dummynet.c and binaries are derived >> from them -- it makes no sense >> that you have 10 pkts in the queue and certainly >> they cannot sum up to over 300Kbytes. >> (it might well be my fault... can you refresh me on your > FreeBSD 3.1-RELEASE, > Here are the files listed I updatet: > bridge.c v 1.6 Thu Feb 25 10:48:31 1999 UTC by kato > if_ed.c $Id: if_ed.c,v 1.150 1999/03/17 16:44:51 luigi Exp $ > ip_dummynet.c $Id: ip_dummynet.c,v 1.8 1999/01/27 22:42:24 dillon Exp $ > ip_fw.c $Id: ip_fw.c,v 1.104 1999/02/16 10:49:52 dfr Exp $ > Mybe I mixed some versions? > >> os and the version of the above files... in doubt rm kernel bridge.o >> ipfw.o ip_dummynet.o from your compile dir and rebuild one.) > I made allways a "make clean" in my compile directory, after I installed > the new sources. > > what does the fields in this record exactly mean? > 00001: 1.000 Mbit/s 0 ms 100 sl. -- 10 pkts (352720 B) 0 drops > | | | | | > pipenr bandwidth delay ?? ???? > > Luigi: Do you read mails send to ? > > Bye, bye, > -- Actualy, your problem have been fixed in the 2.2 Branch. Just wait for a few day for having Luigi commit it in the 3.0 branch. (You will have to get the latest stable). RN. IhM To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 6:21: 5 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id CE16E14C37 for ; Fri, 26 Mar 1999 06:20:56 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id NAA15194; Fri, 26 Mar 1999 13:06:39 +0100 From: Luigi Rizzo Message-Id: <199903261206.NAA15194@labinfo.iet.unipi.it> Subject: Re: Dummynet To: remy@synx.com Date: Fri, 26 Mar 1999 13:06:38 +0100 (MET) Cc: mbretter@sbox.tu-graz.ac.at, freebsd-net@FreeBSD.ORG In-Reply-To: <199903261419.PAA12462@rt2.synx.com> from "Remy Nonnenmacher" at Mar 26, 99 03:11:16 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 237 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Actualy, your problem have been fixed in the 2.2 Branch. Just wait for > a few day for having Luigi commit it in the 3.0 branch. (You will have done right now in -current and -stable. Friday is a good day for fixing things... luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 6:56:48 1999 Delivered-To: freebsd-net@freebsd.org Received: from ausmail1.austin.ibm.com (ausmail1.austin.ibm.com [192.35.232.12]) by hub.freebsd.org (Postfix) with ESMTP id 70A9515508 for ; Fri, 26 Mar 1999 06:56:46 -0800 (PST) (envelope-from marquard@austin.ibm.com) Received: from netmail.austin.ibm.com (netmail.austin.ibm.com [9.53.250.98]) by ausmail1.austin.ibm.com (8.9.1/8.8.5) with ESMTP id IAA30138 for ; Fri, 26 Mar 1999 08:51:20 -0600 Received: from mojave.austin.ibm.com (mojave.austin.ibm.com [9.53.150.76]) by netmail.austin.ibm.com (8.8.5/8.8.5) with ESMTP id IAA56990 for ; Fri, 26 Mar 1999 08:56:27 -0600 Received: (from marquard@localhost) by mojave.austin.ibm.com (AIX4.3/UCB 8.8.8/8.7-client1.01) id IAA33876; Fri, 26 Mar 1999 08:56:25 -0600 To: freebsd-net@freebsd.org Subject: Re: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) References: <199903260054.QAA22060@biggusdiskus.flyingfox.com> From: Dave Marquardt Date: 26 Mar 1999 08:56:25 -0600 In-Reply-To: Jim Shankland's message of "Thu, 25 Mar 1999 16:54:16 -0800 (PST)" Message-ID: Lines: 15 X-Mailer: Gnus v5.6.2/Emacs 19.34 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Jim Shankland writes: > A thought related to this discussion: does it make sense to allow the > send buffers to be larger than the peer's advertised window size? > In other words, why "preposition" those bytes in the kernel before > the peer has indicated a willingness to accept them? Yes, it absolutely makes sense. Let's say you send everything the receiver allows you to send. If your socket send buffer has unsent data in it, then when the receiver ACKs, TCP can start sending right away. If you don't have more data in the socket send buffer, you have to wake up the application and copy more data in from it. So, in other words, having a larger send buffer allows streaming. -Dave To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 7: 4:59 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 66A5C1506E for ; Fri, 26 Mar 1999 07:04:53 -0800 (PST) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.1/8.9.1) id KAA23189; Fri, 26 Mar 1999 10:04:22 -0500 (EST) (envelope-from wollman) Date: Fri, 26 Mar 1999 10:04:22 -0500 (EST) From: Garrett Wollman Message-Id: <199903261504.KAA23189@khavrinen.lcs.mit.edu> To: Jim Shankland Cc: j@lumiere.net, freebsd-net@FreeBSD.ORG Subject: mbuf clusters and socket send buffers (was Re: 3.1-STABLE dies on 40+ connects) In-Reply-To: <199903260054.QAA22060@biggusdiskus.flyingfox.com> References: <199903260054.QAA22060@biggusdiskus.flyingfox.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > A thought related to this discussion: does it make sense to allow the > send buffers to be larger than the peer's advertised window size? > In other words, why "preposition" those bytes in the kernel before > the peer has indicated a willingness to accept them? No, it doesn't. Just ask Van Jacobson. Unfortunately, that doesn't get OUR TCP stack rewritten any faster. -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 Fri Mar 26 7:34:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id E636B15131 for ; Fri, 26 Mar 1999 07:34:17 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id QAA18795 for ; Fri, 26 Mar 1999 16:33:57 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: RE: Dummynet Date: Fri, 26 Mar 1999 16:34:14 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99032616361301.00391@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, now I upgradet ip_dummynet.c to $Id: ip_dummynet.c,v 1.7.2.2 1999/03/26 14:19:40 luigi Exp. Now I got these errors in kernel-compilation: ../../netinet/ip_dummynet.c: In function `dn_move': ../../netinet/ip_dummynet.c:160: `ip' undeclared (first use this function) ../../netinet/ip_dummynet.c:160: (Each undeclared identifier is reported only once ../../netinet/ip_dummynet.c:160: for each function it appears in.) *** Error code 1 Stop. Are ther other source-files to upgrade? thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 8:17:46 1999 Delivered-To: freebsd-net@freebsd.org Received: from maxwell.syr.edu (maxwell.syr.edu [128.230.129.5]) by hub.freebsd.org (Postfix) with ESMTP id B886915657 for ; Fri, 26 Mar 1999 08:17:44 -0800 (PST) (envelope-from cmsedore@maxwell.syr.edu) Received: from exchange.maxwell.syr.edu (exchange.maxwell.syr.edu [128.230.129.241]) by maxwell.syr.edu (8.9.1a/8.9.1) with ESMTP id LAA00895 for ; Fri, 26 Mar 1999 11:04:26 GMT Received: by exchange.maxwell.syr.edu with Internet Mail Service (5.5.1960.3) id ; Fri, 26 Mar 1999 11:17:20 -0500 Message-ID: <262C3DA9BE0CD211971700A0C9B413A1CBEF@exchange.maxwell.syr.edu> From: Christopher Sedore To: "'freebsd-net@freebsd.org'" Subject: bpf bug? Date: Fri, 26 Mar 1999 11:17:17 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.1960.3) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I noticed today that when sending a packet through the bpf interface, the kernel silently resets the ethernet source MAC to the MAC of the network card doing the send. My impression was that bpf writes should go direct to the interface without any edits in the kernel--am I wrong on this? -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 8:49:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5]) by hub.freebsd.org (Postfix) with SMTP id 40DC515714 for ; Fri, 26 Mar 1999 08:49:31 -0800 (PST) (envelope-from luigi@labinfo.iet.unipi.it) Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id PAA00170; Fri, 26 Mar 1999 15:27:04 +0100 From: Luigi Rizzo Message-Id: <199903261427.PAA00170@labinfo.iet.unipi.it> Subject: Re: Dummynet To: mbretter@sbox.tu-graz.ac.at (Michael Bretterklieber) Date: Fri, 26 Mar 1999 15:27:04 +0100 (MET) Cc: freebsd-net@FreeBSD.ORG In-Reply-To: <99032616361301.00391@babsi.tu-graz.ac.at> from "Michael Bretterklieber" at Mar 26, 99 04:33:55 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 581 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Hi, > > now I upgradet ip_dummynet.c to $Id: ip_dummynet.c,v 1.7.2.2 1999/03/26 > 14:19:40 luigi Exp. > > Now I got these errors in kernel-compilation: > > ../../netinet/ip_dummynet.c: In function `dn_move': > ../../netinet/ip_dummynet.c:160: `ip' undeclared (first use this function) > ../../netinet/ip_dummynet.c:160: (Each undeclared identifier is reported only once > ../../netinet/ip_dummynet.c:160: for each function it appears in.) > *** Error code 1 the line should read pipe->r_len_bytes -= len ; i think i forgot to patch this one :( luigi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 9: 5:10 1999 Delivered-To: freebsd-net@freebsd.org Received: from ns1.tu-graz.ac.at (ns1.tu-graz.ac.at [129.27.2.3]) by hub.freebsd.org (Postfix) with ESMTP id 5E06B14C08 for ; Fri, 26 Mar 1999 09:05:00 -0800 (PST) (envelope-from mbretter@sbox.tu-graz.ac.at) Received: from babsi.tu-graz.ac.at (teleweb-17.vc-graz.ac.at [193.171.247.17]) by ns1.tu-graz.ac.at (8.8.6/8.8.6) with SMTP id SAA20106 for ; Fri, 26 Mar 1999 18:04:40 +0100 (MET) From: Michael Bretterklieber To: freebsd-net@freebsd.org Subject: Re: Dummynet Date: Fri, 26 Mar 1999 17:55:30 +0100 X-Mailer: KMail [version 1.0.17] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99032618065700.03931@babsi.tu-graz.ac.at> Content-Transfer-Encoding: 8bit X-KMail-Mark: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, >the line should read > > pipe->r_len_bytes -= len ; > >i think i forgot to patch this one :( I added this and now it works. I also tested bridge and dummynet successfully with my configuration. :) My custom-picobsd with this configuration is now available at: http://www.sbox.tu-graz.ac.at/home/m/mbretter/picobsd.bin Thanx, -- \|/ @ @ +---------------------------oOO-(_)-OOo---------------------------+ Michael Bretterklieber Office: Michael.Bretterklieber@gamed.com Privat: mbretter@sbox.tu-graz.ac.at URL: http://www.sbox.tu-graz.ac.at/home/m/mbretter/ +-----------------------------------------------------------------+ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 10:20:14 1999 Delivered-To: freebsd-net@freebsd.org Received: from bubba.whistle.com (s205m7.whistle.com [207.76.205.7]) by hub.freebsd.org (Postfix) with ESMTP id B27A714C08 for ; Fri, 26 Mar 1999 10:20:13 -0800 (PST) (envelope-from archie@whistle.com) Received: (from archie@localhost) by bubba.whistle.com (8.9.2/8.9.2) id KAA86181; Fri, 26 Mar 1999 10:19:22 -0800 (PST) From: Archie Cobbs Message-Id: <199903261819.KAA86181@bubba.whistle.com> Subject: Re: bpf bug? In-Reply-To: <262C3DA9BE0CD211971700A0C9B413A1CBEF@exchange.maxwell.syr.edu> from Christopher Sedore at "Mar 26, 99 11:17:17 am" To: cmsedore@maxwell.syr.edu (Christopher Sedore) Date: Fri, 26 Mar 1999 10:19:22 -0800 (PST) Cc: freebsd-net@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Christopher Sedore writes: > I noticed today that when sending a packet through the bpf interface, > the kernel silently resets the ethernet source MAC to the MAC of the > network card doing the send. My impression was that bpf writes should > go direct to the interface without any edits in the kernel--am I wrong > on this? My opinion on this is that yes it's a bug. However, it's not so easy to fix, because other parts of the kernel surely rely on this behavior. To fix it would probably require a new argument to if_output() (or whatever it's called) or perhaps a new mbuf flag. I suggest you file a PR and see what comes back :-) -Archie ___________________________________________________________________________ Archie Cobbs * Whistle Communications, Inc. * http://www.whistle.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 10:36:57 1999 Delivered-To: freebsd-net@freebsd.org Received: from abused.com (abused.com [204.216.142.63]) by hub.freebsd.org (Postfix) with ESMTP id CA3D01555F for ; Fri, 26 Mar 1999 10:36:55 -0800 (PST) (envelope-from gvbmail@tns.net) Received: from gvb (gvb.tns.net [204.216.245.137]) by abused.com (8.9.3/I feel abused.) with SMTP id KAA74130 for ; Fri, 26 Mar 1999 10:36:36 -0800 (PST) Message-Id: <4.1.19990326103409.02bf7f00@abused.com> X-Sender: gvbmail@mail.tns.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Fri, 26 Mar 1999 10:36:35 -0800 To: freebsd-net@freebsd.org From: GVB Subject: autoresponder for email.. 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 trying to find a decent auto responder package for a FreeBSD serverside sendmail/qpopper setup. I had looked into using the vacation program which works fine, EXCEPT that it requires users to have a shell to run it. I am not going to start giving shells out on a mail server just for this application.. does anyone know if there is ANYTHING else out there that will fit what I am doing without having to use a new mail server?? Thanks.. G V B To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 11:22:52 1999 Delivered-To: freebsd-net@freebsd.org Received: from ryouko.nas.nasa.gov (ryouko.nas.nasa.gov [129.99.34.113]) by hub.freebsd.org (Postfix) with ESMTP id BB09014D75 for ; Fri, 26 Mar 1999 11:22:46 -0800 (PST) (envelope-from greg@ryouko.nas.nasa.gov) Received: from ryouko.nas.nasa.gov (localhost [127.0.0.1]) by ryouko.nas.nasa.gov (8.8.7/NAS8.8.7n) with ESMTP id LAA16350; Fri, 26 Mar 1999 11:22:09 -0800 (PST) Message-Id: <199903261922.LAA16350@ryouko.nas.nasa.gov> To: Archie Cobbs Cc: cmsedore@maxwell.syr.edu (Christopher Sedore), freebsd-net@FreeBSD.ORG Subject: Re: bpf bug? In-reply-to: Your message of "Fri, 26 Mar 1999 10:19:22 PST." <199903261819.KAA86181@bubba.whistle.com> Date: Fri, 26 Mar 1999 11:22:09 -0800 From: "Gregory P. Smith" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Christopher Sedore writes: > > I noticed today that when sending a packet through the bpf interface, > > the kernel silently resets the ethernet source MAC to the MAC of the > > network card doing the send. My impression was that bpf writes should > > go direct to the interface without any edits in the kernel--am I wrong > > on this? > > My opinion on this is that yes it's a bug. However, it's not so > easy to fix, because other parts of the kernel surely rely on > this behavior. To fix it would probably require a new argument > to if_output() (or whatever it's called) or perhaps a new mbuf flag. > I suggest you file a PR and see what comes back :-) I made a patch for NetBSD 1.3 about a year ago that allowed you to stop this behavior using an ioctl as follows on your bpf file descriptor: ioctl(fd, BIOCSHDRCMPLT, (caddr_t) & one) I believe it got checked into the main NetBSD source tree by (thorpej@nas.nasa.gov) but if you don't see it there let me know and I'll can send the patch which shouldn't be too hard to fit into FreeBSD. Greg To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 13:37:49 1999 Delivered-To: freebsd-net@freebsd.org Received: from carp.gbr.epa.gov (carp.gbr.epa.gov [204.46.159.110]) by hub.freebsd.org (Postfix) with ESMTP id 2410315060 for ; Fri, 26 Mar 1999 13:37:25 -0800 (PST) (envelope-from mjenkins@carp.gbr.epa.gov) Received: (from mjenkins@localhost) by carp.gbr.epa.gov (8.8.8/8.8.8) id PAA06872; Fri, 26 Mar 1999 15:37:01 -0600 (CST) (envelope-from mjenkins) Date: Fri, 26 Mar 1999 15:37:01 -0600 (CST) From: Mike Jenkins Message-Id: <199903262137.PAA06872@carp.gbr.epa.gov> To: mm@i.cz Subject: Re: switch vs bridge (fwd) Cc: freebsd-net@freebsd.org In-Reply-To: Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, 26 Mar 1999 Martin Machacek wrote: > Layer 4 switch is a pure marketing bullshit. If I understand layer 4 switches correctly, they switch at the tcp/udp port number layer. I could therefore slip a layer 4 switch between my router and my lan, and program it to redirect all incoming 25/tcp smtp connections to a mail filter host. I find that rather useful. I'm sure some folks use them for 80/tcp http redirection for web caching. Aren't these useful applicatons? I realize routers can be programmed to do this but who wants to load down (or misconfigure) the router for this chore. A dual-homed unix box such as FreeBSD can also do this using redirection in packet filtering but that usually requires splitting the network into 2 IP networks (yes i've heard of dummynet/bridge but that is work in progress). I think a network appliance like a layer 4 switch would be the right tool for the job. Mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 14: 4:11 1999 Delivered-To: freebsd-net@freebsd.org Received: from us.net (laurel.us.net [198.240.72.4]) by hub.freebsd.org (Postfix) with ESMTP id EBFE6150BB for ; Fri, 26 Mar 1999 14:03:34 -0800 (PST) (envelope-from dgs@us.net) Received: (from dgs@localhost) by us.net (8.8.8/8.8.8) id RAA26464; Fri, 26 Mar 1999 17:02:42 -0500 (EST) X-Provider: US Net - Advanced Internet Services - 301-361-USNET - info@us.net Where Business Connects! (tm) -- http://www.us.net/ From: David Stoddard Message-Id: <199903262202.RAA26464@us.net> Subject: Re: autoresponder for email.. To: gvbmail@tns.net (GVB) Date: Fri, 26 Mar 1999 17:02:40 -0500 (EST) Cc: freebsd-net@freebsd.org In-Reply-To: <4.1.19990326103409.02bf7f00@abused.com> from "GVB" at Mar 26, 99 10:36:35 am X-Mailer: ELM [version 2.4 PL25] Content-Type: text Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org GVB writes: > I am trying to find a decent auto responder package for a FreeBSD > serverside sendmail/qpopper setup. I had looked into using the vacation > program which works fine, EXCEPT that it requires users to have a shell to > run it. I am not going to start giving shells out on a mail server just > for this application.. does anyone know if there is ANYTHING else out there > that will fit what I am doing without having to use a new mail server?? I have one that I wrote that you can use. You will need three things: and alias in your mail system (or .forward file) to send mail to an executable program, the C source for the executable program, and a perl routine that performs the autoresponse function. This is what your alias should look like (assuming you want an info responder): info: |"/full/path/here/autorespond sales@your.domain" --- Next, the autorespond program is a C wrapper that actually executes a perl program. Be sure to modify the path names in the code below to suit your own installation. The C wrapper looks like this: /* C program to wrap /usr/home/info/autorespond.pl */ #include char ifs[] = "IFS= \t\n"; char path[] = "PATH=/bin:/usr/bin:/usr/local/bin:/sbin:/usr/sbin"; char shell[] = "/usr/local/bin/perl"; int main(argc,argv) int argc; char *argv[]; { char **p, **pp; putenv(ifs); putenv(path); setgid(getegid()); setuid(geteuid()); p = pp = (char **) malloc(sizeof(char*) * (argc + 2)); if (!p) { fprintf(stderr, "mkwrap: %s: Cannot malloc parmlist\n",argv[0]); return 6; } *p++ = (argc--, *argv++); *p++ = "/usr/home/info/autorespond.pl"; while (argc > 0) *p++ = (argc--, *argv++); *p = NULL; execv(shell,pp); perror(shell); return 8; } Compile this program on FreeBSD using: cc autorespond.c -o autorespond The program should be flagged via "chmod 755 autorespond". Note that sendmail requires perl routines to be wrapped by natively executable binary programs, hence the reason for the wrapper. --- Finally, the actual autoresponder is included below. This is one of several I have written over the years and is the result of much frustration I have encountered with broken mail programs all over the planet. It works well, avoids mail loops, logs each person that requests info from it, and is very configurable. It even provides a filter mechanism to stop people that abuse it as a test facility. Modify the path names in the code to suit your installation. The perl routine is attached. Dave Stoddard US Net Incorporated 301-361-6000 dgs@us.net --- #!/usr/local/bin/perl # autorespond.pl - an email autoresponder written in perl. # # Dave Stoddard # US Net Incorporated # dgs@us.net # # Legal Disclaimer # ---------------- # Permission to use and distribute this program is given provided the # original author of this program is given credit for his work. This # program is not guaranteed to do anything other than consume space on # your disk drive. Use it at your own risk. # # Description # ----------- # This program is designed to read an email message from stdin, determine # who sent it, and mail a response back to the sender. The person that # sent the message is logged to a log file. Also, a filter file is available # to block people who abuse the autoresponder. The name of the file to # send (the response file), the log file name, and the filter file name # are all defined in the beginning of this script and should be tailored # for each specific installation. # # The program can also be run completely from parameters, allowing someone # to use it for multiple purposes on a single site. The parameters are # optional and are defined as follows: # # autorespond.pl [ -m message ] [ -l log ] [ -f filter ] [ -s -e ] [ email ] # # where, # # -m message : specifies the message file to send # -l log : specifies the name of the file used to log recipients # -f filter : specifies the name of the filter file used to block abusers # -e : suppress error messages from the autoresponder # -s : preserve the Subject: header in the message file # email : a list of email recipients for the original input message # # It is possible to specify any combination of parameters and let other # parameters default to internal values. By specifying one or more email # addresses at the end of the line, you can send a copy of the person's # message to one or more email addresses. This is useful for answering # specific questions that may be sent to the autoresponder (common for # sales information). # # It is important to note that the "message.file" should contain any necessary # mail headers in the beginning of the file, followed by a blank line, then # followed by the message you want to send. The mail headers will allow you # to control the subject of the message, the reply-to address, and other # elements of the message. This is a typical example for a message file: # # Subject: Information You Requested # Reply-To: sales@us.net # Errors-To: postmaster@us.net # # This is a test message. # # Note that all of the lines above should be left justified in the text file. # If you use a mail transport agent other than sendmail, such as smail or # MMDF, this technique of embedding the message headers in the message file # may not work for you -- you will need to test this on your own. A complete # compendium of email message headers is available in the Internet standards # document RFC-822, available at http://www.internic.net/rfc/rfc822.txt (and # other locations too). # # IMPORTANT MODIFICATION: Because some autoresponders use a subject line # to for loop detection, if the original message has a subject line, the # autoresponder will preserve that subject line and override the subject # line you specify in the message file. This action can be overridden # using the -s option on the command line. # # Also, it may be necessary to add a definition in your mail server to allow # the mail server to run this program. In sendmail.cf, this is called a # "trusted user" and is defined with the "Tuserid" macro. You should make # sure your log file can be written to by your mail server. If you know # the userid and group your mail server run under, you can set the ownership # of the file accordingly and use chmod to enable write access. If you don't # know how to determine this, then simply "chmod 666" your log file -- if you # do this, other users on your system will be able to access your log file as # well. Your best bet is to set it correctly than to use "666" on the file. # # One last point -- all files referenced by autorespond.pl must be world # readable. This is done as a security precaution. Don't remove this # security code, or you run the risk of having any file on your system # vulnerable to being mailed. # # Example # ------- # Assuming you have an email alias called "info" that you use for sending # information packets, and you want a copy of the original request to be # sent to the address "sales" (in case the request has questions), your # mail alias definition would look like this: # # info: "|/usr/local/bin/autorespond -m /etc/myinfo -l /etc/mylog.txt sales" # # Mail arriving for "info" would receive a response using the contents of # the file /etc/myinfo, and the response would be logged to the file # /etc/mylog.txt. Note that you could do the same thing with a .forward # file that would contain the following entries: # # |/usr/local/bin/autorespond -m /etc/myinfo -l /etc/mylog.txt sales # # An alternative .forward approach is to put the recipient userid outside # the autoresponder, however filtering will not be in effect for any email # address that is outside the command line for the autoresponder: # # |/usr/local/bin/autorespond -m /etc/myinfo -l /etc/mylog.txt # sales # # If you want to simply accept the defaults coded into the script, all # you need to refer to is the /usr/local/bin/autorespond program itself. # In this case, the following alias would work just fine (assuming filter # for recipient copy address): # # info: "|/usr/local/bin/autorespond sales" # # or without filter for recipient copy address: # # info: "|/usr/local/bin/autorespond",sales # # This would send the default response file, log to the default log file, # apply default filters, and send a copy of the message to "sales". # # Components # ---------- # Because this routine is a perl script, it needs to be executed by a # wrapper program that is written in C. You will need to examine the # program autorespond.c and add the correct path info for your perl # interpreter on your system, and the path of the autorespond.pl script. # To compile sutorespond.c, simply use: # # cc autorespond.c -o autorespond # # The autorespond program simply invokes autorespond.pl using the perl # interpreter. The important elements you will want to modify in this # program include: # # $errmail : email address to use to send responder error messages # $sendfile : the default message file to send # $logfile : the default log file to use for logging recipients # $filter : the default filter file to use # # Note that if $errmail, $logfile, or $filter are blank, they will be # ignored by the software (the functionality os removed). # # Revision History # ---------------- # 1.0 - 02/16/98 - d. stoddard - initial script passed final testing. # 1.1 - 06/22/98 - d. stoddard - added loop control for error report. # 2.0 - 02/16/98 - d. stoddard - added senders headers, internal copying # of message to recipients, and option switches. # require "getopts.pl"; $errmail = "support\@your.domain"; # where to send errors $sendfile = "/usr/home/info/autorespond.txt"; # the default file to send $logfile = "/usr/home/info/autorespond.log"; # the log file for each request $filter = "/usr/home/info/autorespond.filter"; # addresses to block $tmpfile = "/tmp/autorespond.$$"; # temporary file $inbody = 0; # email message body flag # UNIX commands $sendmail = "/usr/sbin/sendmail -odi"; # the path to sendmail $datecmd = "/bin/date"; # the path to the date command ### ### Step 1: Process command line parameters. ### # -m message : specifies the message file to send # -l log : specifies the name of the file used to log recipients # -f filter : specifies the name of the filter file used to block abusers # -e : suppress autoresponder email error reporting # -s : preserve the Subject: header in the message file # email : a list of email recipients for the original input message &Getopts ('def:l:m:s'); # debugging option if ($opt_d) { $debug = 1; } # error suppression option if ($opt_e) { $noerrout = 1; } # preserve message file subject option if ($opt_s) { $keepsubj = 1; } # if they specified a filter file, use it if ($opt_f) { $filter = $opt_f; } # if they specified a log file, use it if ($opt_l) { $logfile = $opt_l; } # if they specified a response file, use it if ($opt_m) { $sendfile = $opt_m; } # save any optional email recipients @emailrcp = @ARGV; # format a date stamp for the log file $date = `$datecmd "+%D %T"`; # get a date stamp ### ### Step 2: Read through the message header and find the return address. ### *** This saves the incoming message to a temporary file *** ### # open stdin open (INP, "-") || &HandleError ("Error $|: can not open stdin"); open (TMP, ">$tmpfile") || &HandleError ("Error $|: can not open $tmpfile"); while () { # save the input lines print TMP "$_"; # look for our loop control element ANYWHERE in the message if (/X-Loop-Control: Autoresponder/) { $errmsg = "Error: Message loop detected!"; $error = 1; } # detect the change from message header to message body if ($_ eq "\n") { $inbody = 1; } # if we are in the message body, get another line $inbody && next; # handle From (from-space) if (/^from \s*(\S+)\s*/i) { $fromspace = $1; next; } # handle From: if (/^from: \s*(\S+)\s*/i) { $from = $1; # if it is embedded in angle brackets, strip it out if (/\<(\S+)\>/) { $from = $1; next; } if (/\s+(\S+)\@(\S+)\s*/) { $from = $1 . "\@" . $2; next; } } # handle Reply-To: if (/^reply-to: \s*(\S+)\s*/i) { $replyto = $1; # if it is embedded in angle brackets, strip it out if (/\<(\S+)\>/) { $replyto = $1; next; } if (/\s+(\S+)\@(\S+)\s*/) { $replyto = $1 . "\@" . $2; next; } } # # Code added here to save most of the senders header (for loop control). # # save the subject line if (/^subject:(.*)$/i) { if ($keepsubj) { # check for subject override next; } if (/^subject:\s*$/i) { # blank subject line -- ignore next; } push (@extrahdrs,"Subject: Re:$1"); $hassubject = 1; next; } # dump received headers if (/^received:/i) { next; } # handle continuation lines if (/^\s+/) { next; } # dump provider headers if (/^x-provider:/i) { next; } # dump organization headers if (/^organization:/i) { next; } # dump message-id headers if (/^message-id:/i) { next; } # dump to headers if (/^to:/i) { next; } # dump apparently-to headers if (/^apparently-to:/i) { next; } # dump date headers if (/^date:/i) { next; } # dump x-mailer headers if (/^x-mailer:/i) { next; } # dump content-type headers if (/^content-type:/i) { next; } chop; push (@extrahdrs,$_); } close (INP); close (TMP); # if we encountered an error during processing, blast out of the message if ($error) { &HandleError ($errmsg); } ### ### Step 3: Check file attributes to avoid security holes. ### *** Done after the original message is saved to temp file *** ### # stat the message reply file and retrieve its permissions ($dev,$inode,$mode,$nlink,$uid,$gid,$rdev, $size,$atime,$mtime,$ctime,$blksz,$blocks) = stat ($sendfile); # convert the mode into a binary value $val = pack ("I",$mode); # convert the mode bits into a bit array foreach $i (0 .. 31) { @bits[$i] = vec ($val,$i,1); } $othx = $bits[0]; # other execute $othw = $bits[1]; # other write $othr = $bits[2]; # other read $grpx = $bits[3]; # group execute $grpw = $bits[4]; # group write $grpr = $bits[5]; # group read $usrx = $bits[6]; # owner execute $usrw = $bits[7]; # owner write $usrr = $bits[8]; # owner read $sett = $bits[9]; # sticky bit $setg = $bits[10]; # set gid bit $setu = $bits[11]; # set uid bit # if the file is not world readable, head south ... if ($othr != 1) { &HandleError ("Permission Error: $sendfile is not world readable"); # never returns ... } # uncomment if you want to prevent files owned by root from download #if ($uid == 0) { # &HandleError ("Permission Error: $sendfile is owned by root"); # # never returns ... #} #if ($logfile ne "") { # ($dev,$inode,$mode,$nlink,$uid,$gid) = stat ($logfile); # if ($uid == 0) { # &HandleError ("Permission Error: $logfile is owned by root"); # # never returns ... # } #} #if ($filter ne "") { # ($dev,$inode,$mode,$nlink,$uid,$gid) = stat ($filter); # if ($uid == 0) { # &HandleError ("Permission Error: $filter is owned by root"); # # never returns ... # } #} ### ### Step 4: Determine which email return address to use. ### if ("$replyto" ne "") { $dest = $replyto; } elsif ("$from" ne "") { $dest = $from; } elsif ("$fromspace" ne "") { $dest = $fromspace; } else { &HandleError("Unable to determine originator of message"); } ### ### Step 5: Don't respond to postmaster, mailer-daemon, daemon, or maiser ### in order to avoid loops. ### $dest =~ y/A-Z/a-z/; ($userid,$junk) = split (/\@/,$dest); if ("$userid" eq "postmaster") { &HandleError("Error: Message originator is postmaster\!"); } elsif ("$userid" eq "mailer-daemon") { &HandleError("Error: Message originator is mailer-daemon\!"); } elsif ("$userid" eq "daemon") { &HandleError("Error: Message originator is daemon\!"); } elsif ("$fromspace" eq "daemon") { &HandleError("Error: Message originator is daemon\!"); } elsif ("$userid" eq "maiser") { &HandleError("Error: Message originator is maiser\!"); } ### ### Step 6: If we have a filter file, apply the filter rules ### if ($filter ne "") { open (FLT, "$filter") || &HandleError ("Error $|: can not open $filter"); while () { chop; next if /^$/; # dump blank lines next if /^[\s]*[\#]/; # dump comments if ($dest eq $_) { &LogEntry ("blocked by filter"); close (FLT); unlink ($tmpfile); exit 0; } } close (FLT); } ### ### Step 7: Send the response file back to the originator of the request. ### open (MSG,"$sendfile") || &HandleError("error $|: can not open $sendfile"); open (OUT,"| $sendmail $dest") || &HandleError("error $|: invoking sendmail"); # check for option to preserve message file subject line if ($keepsubj) { $hassubject = 0; } # drop in our loop control header print OUT "X-Loop-Control: Autoresponder\n"; # print the additional user headers (for the sender's loop control) foreach $header (@extrahdrs) { print OUT "$header\n"; } while () { if ($hassubject == 1) { if (/^Subject: /) { next; } } print OUT "$_"; } close MSG; close OUT; ### ### Step 8: If email recipients exist, send a copy of the message to them. ### if (@emailrcp) { open (TMP,"$tmpfile") || &HandleError("error $|: can not open $tmpfile"); open (OUT,"| $sendmail @emailrcp") || &HandleError("error $|: invoking sendmail"); while () { print OUT "$_"; } close TMP; close OUT; } ### ### Step 9: Log the request to the log file. ### &LogEntry (); unlink ($tmpfile); exit 0; ### END MAIN PROGRAM ### ### ### Generic Error Handler ### sub HandleError { local ($error) = @_; # if they disabled error reporting, just get out if ($noerrout == 1) { unlink ($tmpfile); &LogEntry ($error); exit 0; } if ($errmail ne "") { open (OUT,"| $sendmail $errmail"); print OUT <) { print OUT "$_"; } close (TMP); close (OUT); } unlink ($tmpfile); &LogEntry ($error); exit 0; } sub LogEntry { local ($msg) = @_; if ($logfile ne "") { open (LOG,">>$logfile") || print "error $|: can not open $logfile\n"; if ($msg ne "") { print LOG "$date $dest $sendfile : $msg\n"; } else { print LOG "$date $dest $sendfile\n"; } close (LOG); } } To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 15:47:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from sydney.alpha.net.au (sydney.alpha.net.au [203.31.171.20]) by hub.freebsd.org (Postfix) with ESMTP id EB39214EC2 for ; Fri, 26 Mar 1999 15:47:41 -0800 (PST) (envelope-from danny@alpha.net.au) Received: from fluffy (p11-nas8.syd.ihug.com.au [206.17.113.75]) by sydney.alpha.net.au (8.8.5/8.8.5) with ESMTP id JAA27392; Sat, 27 Mar 1999 09:58:30 +1000 (EST) Message-Id: <199903262358.JAA27392@sydney.alpha.net.au> Reply-To: From: "Danny Ho" To: , "Tony Finch" Subject: OSI layering Query.. Please Help ME Date: Sat, 27 Mar 1999 11:05:41 +1100 X-MSMail-Priority: Normal X-Priority: 3 X-Mailer: Microsoft Internet Mail 4.70.1155 MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi everyone I am sort of new( 1 year experience in this ISP job) Anyway, I have one query about OSI layering. What are some of the advantages of OSI layer as seen in the OSI model?? And what exactly are PDU? What exactly is SAP?? I am really confused help me. ############################################################ Everything in the West is secrets unless there is a consious decision to the contrary. Our civlization, which never stop declaiming about the inviolabilty of free speech, but operates as if it distrusts nothing more. Knowledge is the measurement of your power. Everything in the internet is free speech unless there is a decision to the contrary . Connect to the internet NOW join Alpha Dot Net Australia call:- 9211 7782 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 16:41:28 1999 Delivered-To: freebsd-net@freebsd.org Received: from xylan.com (postal.xylan.com [208.8.0.248]) by hub.freebsd.org (Postfix) with ESMTP id B018815170 for ; Fri, 26 Mar 1999 16:41:23 -0800 (PST) (envelope-from wes@softweyr.com) Received: from mailhub.xylan.com by xylan.com (8.8.7/SMI-SVR4 (xylan-mgw 2.2 [OUT])) id QAA29385; Fri, 26 Mar 1999 16:40:12 -0800 (PST) Received: from utah.XYLAN.COM by mailhub.xylan.com (SMI-8.6/SMI-SVR4 (mailhub 2.1 [HUB])) id QAA27974; Fri, 26 Mar 1999 16:40:09 -0800 Received: from softweyr.com by utah.XYLAN.COM (SMI-8.6/SMI-SVR4 (xylan utah [SPOOL])) id RAA14723; Fri, 26 Mar 1999 17:40:06 -0700 Message-ID: <36FC28DA.73DC2E28@softweyr.com> Date: Fri, 26 Mar 1999 17:39:54 -0700 From: Wes Peters Organization: Softweyr LLC X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Mike Jenkins Cc: mm@i.cz, freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) References: <199903262137.PAA06872@carp.gbr.epa.gov> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Mike Jenkins wrote: > > On Fri, 26 Mar 1999 Martin Machacek wrote: > > Layer 4 switch is a pure marketing bullshit. > > If I understand layer 4 switches correctly, they switch > at the tcp/udp port number layer. To a limited extent, yes. Most "layer 4 switches" implement a very limited version of this. > I could therefore slip > a layer 4 switch between my router and my lan, and program If you have a layer 3 switch, you don't need a router. Just put a wide-area "blade" in the switch and route there. *Good* switches router much faster than routers anyhow. I can't tell you how much faster right now, or I'd have to kill you, but it's MUCH faster. ;^) > it to redirect all incoming 25/tcp smtp connections to a > mail filter host. I supposed you could do that. It's usually used the other way around, to try to provide a crude form of load balancing across mutiple http (i.e.) servers. This turns out to be about as effective as round-robin DNS; a true load balancer would be much more effective. > I find that rather useful. I'm sure > some folks use them for 80/tcp http redirection for web > caching. Well, more likely for bandwidth/performance management and hot failover. > Aren't these useful applicatons? Yes, but the actual features of most of these so-called "layer 4 switches" is so minimal that you'll outgrow them almost immediately, at which time you'd be better off with a REAL load balancer and a less expensive but faster layer 3 switch. > I realize routers can be programmed to do this but who wants > to load down (or misconfigure) the router for this chore. The switch *is* the router, unless you've just got balls of money you're aching to get rid of. If so, call me. We can work together on this. ;^) > A dual-homed unix box such as FreeBSD can also do this using > redirection in packet filtering but that usually requires > splitting the network into 2 IP networks (yes i've heard > of dummynet/bridge but that is work in progress). I think > a network appliance like a layer 4 switch would be the right > tool for the job. The you either don't understand the job, or don't understand the (very limited) capabilities of these so-called layer 4 switches. It's not that it's a bad idea, just that there are a couple of vendors out there giving the idea a bad name with their implementations. -- Where am I, and what am I doing in this handbasket? Wes Peters +1.801.915.2061 Softweyr LLC wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 16:57:35 1999 Delivered-To: freebsd-net@freebsd.org Received: from maxwell.syr.edu (maxwell.syr.edu [128.230.129.5]) by hub.freebsd.org (Postfix) with ESMTP id 84BE714E3E for ; Fri, 26 Mar 1999 16:57:31 -0800 (PST) (envelope-from cmsedore@maxwell.syr.edu) Received: from exchange.maxwell.syr.edu (exchange.maxwell.syr.edu [128.230.129.241]) by maxwell.syr.edu (8.9.1a/8.9.1) with ESMTP id TAA10200 for ; Fri, 26 Mar 1999 19:44:07 GMT Received: by exchange.maxwell.syr.edu with Internet Mail Service (5.5.1960.3) id ; Fri, 26 Mar 1999 19:57:02 -0500 Message-ID: <262C3DA9BE0CD211971700A0C9B413A1CBF1@exchange.maxwell.syr.edu> From: Christopher Sedore To: "'freebsd-net@freebsd.org'" Subject: bpf output bug fix Date: Fri, 26 Mar 1999 19:56:53 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.1960.3) Content-Type: text/plain Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Since it appears that 2 people (3 if you count me :) agree that having the kernel overwrite the source ethernet mac address on sends through bpf is a bug, I fixed it (patch on the end). Basically, rather than setting the dst->sa_family to AF_UNSPEC in the bpfwrite call, I set it to AF_LINK instead. I added glue to if_ethersubr.c to handle AF_LINK, implying that this usage would not conflict with anything else since it wasn't handled previously. It works as expected in my testing. -Chris diff -c -r sys/net/bpf.c /usr/src/sys/net/bpf.c *** sys/net/bpf.c Mon Dec 7 21:58:36 1998 --- /usr/src/sys/net/bpf.c Fri Mar 26 19:52:37 1999 *************** *** 177,185 **** break; case DLT_EN10MB: ! sockp->sa_family = AF_UNSPEC; ! /* XXX Would MAXLINKHDR be better? */ ! hlen = sizeof(struct ether_header); break; case DLT_FDDI: --- 177,184 ---- break; case DLT_EN10MB: ! sockp->sa_family = AF_LINK; ! hlen = 0; break; case DLT_FDDI: diff -c -r sys/net/if_ethersubr.c /usr/src/sys/net/if_ethersubr.c *** sys/net/if_ethersubr.c Tue Jan 12 12:07:00 1999 --- /usr/src/sys/net/if_ethersubr.c Fri Mar 26 18:58:58 1999 *************** *** 336,360 **** type = eh->ether_type; break; default: printf("%s%d: can't handle af%d\n", ifp->if_name, ifp->if_unit, dst->sa_family); senderr(EAFNOSUPPORT); } ! /* ! * Add local net header. If no space in first mbuf, ! * allocate another. ! */ ! M_PREPEND(m, sizeof (struct ether_header), M_DONTWAIT); ! if (m == 0) ! senderr(ENOBUFS); ! eh = mtod(m, struct ether_header *); ! (void)memcpy(&eh->ether_type, &type, ! sizeof(eh->ether_type)); ! (void)memcpy(eh->ether_dhost, edst, sizeof (edst)); ! (void)memcpy(eh->ether_shost, ac->ac_enaddr, ! sizeof(eh->ether_shost)); /* * If a simplex interface, and the packet is being sent to our --- 336,369 ---- type = eh->ether_type; break; + case AF_LINK: + loop_copy = -1; + break; + default: printf("%s%d: can't handle af%d\n", ifp->if_name, ifp->if_unit, dst->sa_family); senderr(EAFNOSUPPORT); } ! if (dst->sa_family!=AF_LINK) { ! ! /* ! * Add local net header. If no space in first mbuf, ! * allocate another. ! */ ! M_PREPEND(m, sizeof (struct ether_header), M_DONTWAIT); ! if (m == 0) ! senderr(ENOBUFS); ! eh = mtod(m, struct ether_header *); ! (void)memcpy(&eh->ether_type, &type, ! sizeof(eh->ether_type)); ! (void)memcpy(eh->ether_dhost, edst, sizeof (edst)); ! (void)memcpy(eh->ether_shost, ac->ac_enaddr, ! sizeof(eh->ether_shost)); ! } else { ! eh = mtod(m, struct ether_header *); ! } /* * If a simplex interface, and the packet is being sent to our To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 17: 4:57 1999 Delivered-To: freebsd-net@freebsd.org Received: from sumatra.americantv.com (sumatra.americantv.com [207.170.17.37]) by hub.freebsd.org (Postfix) with ESMTP id 2733114A14 for ; Fri, 26 Mar 1999 17:04:45 -0800 (PST) (envelope-from jlemon@americantv.com) Received: from right.PCS (right.PCS [148.105.10.31]) by sumatra.americantv.com (8.8.5/8.8.5) with ESMTP id TAA26642; Fri, 26 Mar 1999 19:04:25 -0600 (CST) Received: from free.pcs (free.PCS [148.105.10.51]) by right.PCS (8.6.13/8.6.4) with ESMTP id TAA07131; Fri, 26 Mar 1999 19:03:54 -0600 Received: (from jlemon@localhost) by free.pcs (8.8.6/8.8.5) id TAA14685; Fri, 26 Mar 1999 19:03:54 -0600 (CST) Date: Fri, 26 Mar 1999 19:03:54 -0600 (CST) From: Jonathan Lemon Message-Id: <199903270103.TAA14685@free.pcs> To: wes@softweyr.com, freebsd-net@freebsd.org Subject: Re: switch vs bridge (fwd) X-Newsgroups: local.mail.freebsd-net In-Reply-To: References: Organization: Architecture and Operating System Fanatics Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In article you write: >Mike Jenkins wrote: >> I find that rather useful. I'm sure >> some folks use them for 80/tcp http redirection for web >> caching. > >Well, more likely for bandwidth/performance management and hot failover. That's actually what the particular switch I have to deal with is doing; it an Alteon switch, set up as a transparent web proxy server. As I understand it, it intercepts all web traffic from the campus and directs it to a pool of proxies. I suppose it's useful for what it does. How would a layer-3 switch do the same thing? -- Jonathan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 17:34:42 1999 Delivered-To: freebsd-net@freebsd.org Received: from dnai.com (dnai.com [207.181.194.98]) by hub.freebsd.org (Postfix) with ESMTP id 13A8E14F93 for ; Fri, 26 Mar 1999 17:34:36 -0800 (PST) (envelope-from miket@dnai.com) Received: from einstein (dnai-207-181-255-13.dialup.dnai.com [207.181.255.13]) by dnai.com (8.8.8/8.8.8) with SMTP id RAA28878; Fri, 26 Mar 1999 17:34:10 -0800 (PST) Message-Id: <4.1.19990326173236.00a0c890@mail.dnai.com> X-Sender: miket@mail.dnai.com X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Fri, 26 Mar 1999 17:33:24 -0800 To: Tony Finch , net@FreeBSD.ORG From: Mike Thompson Subject: Re: IP alias configuration for high availability In-Reply-To: References: <4.1.19990323134009.0099fe10@mail.dnai.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Tony, Thanks for the clever tip. This was the type of helpful information I was looking for. Mike >You can avoid the need to fiddle with arp and ifconfig by setting up >the alias on the server's loopback interface. You then control which >servers are live by changing the routes on the router; the servers >don't need to be told what's going on. > >To set up a server: > >ifconfig ed0 inet 192.168.0.1 netmask 255.255.255.0 >ifconfig lo0 inet 209.185.152.1 netmask 255.255.255.255 alias > >On the router: > >route add -host 209.185.152.1 192.168.0.1 > >Tony. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 17:43:26 1999 Delivered-To: freebsd-net@freebsd.org Received: from dnai.com (dnai.com [207.181.194.98]) by hub.freebsd.org (Postfix) with ESMTP id D44BE14F10 for ; Fri, 26 Mar 1999 17:43:24 -0800 (PST) (envelope-from miket@dnai.com) Received: from einstein (dnai-207-181-255-38.dialup.dnai.com [207.181.255.38]) by dnai.com (8.8.8/8.8.8) with SMTP id RAA01437 for ; Fri, 26 Mar 1999 17:43:05 -0800 (PST) Message-Id: <4.1.19990326173538.00a46490@mail.dnai.com> X-Sender: miket@mail.dnai.com X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Fri, 26 Mar 1999 17:42:15 -0800 To: freebsd-net@freebsd.org From: Mike Thompson Subject: PPP as a bridge over the Internet Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I've been considering using FreeBSD 2.2.8 with PPP OVER TCP to bridge two remote private networks over the Internet. The following caveat appears in the PPP man page. Does anyone have real-world experience with this being a problem? "The networks are effectively bridged - the underlying TCP connection may be across a public network (such as the Internet), and the PPP traffic is conceptually encapsulated (although not packet by packet) inside the TCP stream between the two gateways. The major DISADVANTAGE of this mechanism is that there are two "guaranteed delivery" mechanisms in place - the underlying TCP stream and whatever protocol is used over the PPP link - probably TCP again. If packets are lost, both levels will get in each others way trying to negotiate sending of the missing packet." Mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 17:49:28 1999 Delivered-To: freebsd-net@freebsd.org Received: from carp.gbr.epa.gov (carp.gbr.epa.gov [204.46.159.110]) by hub.freebsd.org (Postfix) with ESMTP id 9B95D15067 for ; Fri, 26 Mar 1999 17:49:25 -0800 (PST) (envelope-from mjenkins@carp.gbr.epa.gov) Received: (from mjenkins@localhost) by carp.gbr.epa.gov (8.8.8/8.8.8) id TAA07234; Fri, 26 Mar 1999 19:49:01 -0600 (CST) (envelope-from mjenkins) Date: Fri, 26 Mar 1999 19:49:01 -0600 (CST) From: Mike Jenkins Message-Id: <199903270149.TAA07234@carp.gbr.epa.gov> To: freebsd-net@freebsd.org Subject: Re: switch vs bridge (fwd) Cc: mm@i.cz, wes@softweyr.com In-Reply-To: <36FC28DA.73DC2E28@softweyr.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Wes Peters wrote: > The you either don't understand the job, or don't understand the (very > limited) capabilities of these so-called layer 4 switches. It's not that > it's a bad idea, just that there are a couple of vendors out there giving > the idea a bad name with their implementations. I think the concept of switching based on udp/tcp port is useful. One thing I'm not clear on is at what layer is the switch gathering the packets? Is it at the data link (Ethernet) layer or the network layer (IP)? I was assuming the data link layer to allow transparency. If it is the network layer then the switch is your router as Wes said. Do layer 3 switches have the ability to redirect packets based on udp/tcp port numbers? Mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 20: 0:52 1999 Delivered-To: freebsd-net@freebsd.org Received: from magicnet.magicnet.net (magicnet.magicnet.net [204.96.116.9]) by hub.freebsd.org (Postfix) with ESMTP id CEB6F14F93 for ; Fri, 26 Mar 1999 20:00:50 -0800 (PST) (envelope-from bill@bilver.magicnet.net) Received: (from uucp@localhost) by magicnet.magicnet.net (8.8.6/8.8.8) with UUCP id XAA25101 for freebsd-net@freebsd.org; Fri, 26 Mar 1999 23:00:17 -0500 (EST) Received: (from bill@localhost) by bilver.magicnet.net (8.9.1/8.9.1) id WAA45159 for freebsd-net@freebsd.org; Fri, 26 Mar 1999 22:53:30 -0500 (EST) From: Bill Vermillion Message-Id: <199903270353.WAA45159@bilver.magicnet.net> Subject: Re: switch vs bridge (fwd) In-Reply-To: <36FC28DA.73DC2E28@softweyr.com> from Wes Peters at "Mar 26, 1999 5:39:54 pm" To: freebsd-net@freebsd.org Date: Fri, 26 Mar 1999 22:53:29 -0500 (EST) X-Mailer: ELM [version 2.4ME+ PL43 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Wes Peters recently said: > Mike Jenkins wrote: > If you have a layer 3 switch, you don't need a router. A layer three switch is basically just a router configured in hardware as opposed to software. That's part of the claim of wire-speed switching. -- bill@bilver.magicnet.net | bv@wjv.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 23:24:58 1999 Delivered-To: freebsd-net@freebsd.org Received: from allegro.lemis.com (allegro.lemis.com [192.109.197.134]) by hub.freebsd.org (Postfix) with ESMTP id 3098814C8E; Fri, 26 Mar 1999 23:24:53 -0800 (PST) (envelope-from grog@freebie.lemis.com) Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by allegro.lemis.com (8.9.1/8.9.0) with ESMTP id RAA13750; Sat, 27 Mar 1999 17:54:31 +1030 (CST) Received: (from grog@localhost) by freebie.lemis.com (8.9.3/8.9.0) id RAA53305; Sat, 27 Mar 1999 17:47:39 +1030 (CST) Message-ID: <19990327174738.B425@lemis.com> Date: Sat, 27 Mar 1999 17:47:38 +1030 From: Greg Lehey To: Jesse , "Daniel C. Sobral" Cc: freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Taking panic dumps (was: 3.1-STABLE dies on 40+ connects (resolved)) References: <36FACE5B.D0926C2C@newsguy.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.2i In-Reply-To: ; from Jesse on Thu, Mar 25, 1999 at 04:09:44PM -0800 WWW-Home-Page: http://www.lemis.com/~grog Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Thursday, 25 March 1999 at 16:09:44 -0800, Jesse wrote: > >>> Thanks everybody for your help! FreeBSD is great and it's 80% about the >>> community. Not that I didn't always know it. =) >> >> Notice that a core dump + kernel with symbols would still be very >> useful to find *where* in our code is the bug. > > I asked in other messages if anyone still wanted me to do that. No one > replied. Anyway, I'll do this at an off hour and get a core dump. OK. I really should put this in the FAQ, since it's really important. 1. Build a kernel with debug symbols. Assuming your config file is called MYKERNEL, do: n cd /sys/i386/conf config -g MYCONFIG cd ../../compile/MYCONFIG make depend all cp kernel kernel.gdb strip -g kernel make install 2. If you don't have the directory /var/crash, create it in a file system which has space for the dump (memory size) and the kernel (about 10 MB, 15 is safer). 3. You *must* have a single swap area which is at least a little larger than physical memory. You can have other swap areas as well, but for dumping you need one which will take the contents of memory and a little additional for headers (1 MB would be more than enough extra). Assuming you have such a partition on /dev/da0s1b, add a line like this to /etc/rc.conf: dumpdev="/dev/da0s1b" # Device name to crashdump to (if enabled). If you don't have an /etc/rc.conf, put it in /etc/defaults/rc.conf. 4. Boot the new kernel. There are a number of points to comment about: 1. If you have the space (about 10 MB), you don't need to copy kernel to kernel.gdb and strip the original. It doesn't buy you much, though, unless you have ddb installed, and that is rather painful to use. 2. /etc/rc.conf and friends are still moving around and changing their relationships to each other. This applies to 3.1-RELEASE. It doesn't apply exactly to previous or subsequent releases, though you should be able to figure out what's going on. Greg -- See complete headers for address, home page and phone numbers finger grog@lemis.com for PGP public key To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Fri Mar 26 23:44:59 1999 Delivered-To: freebsd-net@freebsd.org Received: from dnai.com (dnai.com [207.181.194.98]) by hub.freebsd.org (Postfix) with ESMTP id 9BC0314F28 for ; Fri, 26 Mar 1999 23:44:58 -0800 (PST) (envelope-from miket@dnai.com) Received: from einstein (dnai-207-181-255-46.dialup.dnai.com [207.181.255.46]) by dnai.com (8.8.8/8.8.8) with SMTP id XAA09253 for ; Fri, 26 Mar 1999 23:44:36 -0800 (PST) Message-Id: <4.1.19990326233717.009c8210@mail.dnai.com> X-Sender: miket@mail.dnai.com X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Fri, 26 Mar 1999 23:43:42 -0800 To: freebsd-net@freebsd.org From: Mike Thompson Subject: FreeBSD as a router Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org This is probably a stupid question (not that it ever stopped me before), but is a FreeBSD system capable of running as a high-capacity Internet router? I would like to run a number of FreeBSD servers running a web application behind a FreeBSD system acting as a router/firewall. Any examples of web sites doing this would be great. Mike Thompson To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 3:18:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from lion.butya.kz (butya-gw.butya.kz [194.87.112.252]) by hub.freebsd.org (Postfix) with ESMTP id 867F2150C9 for ; Sat, 27 Mar 1999 03:18:31 -0800 (PST) (envelope-from bp@butya.kz) Received: from bp (helo=localhost) by lion.butya.kz with local-esmtp (Exim 2.12 #1) id 10Qr6O-0000Hy-00 for freebsd-net@freebsd.org; Sat, 27 Mar 1999 17:18:08 +0600 Date: Sat, 27 Mar 1999 17:18:08 +0600 (ALMT) From: Boris Popov To: freebsd-net@freebsd.org Subject: NetWare client 1.3beta6 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 Hello, New version is out. It includes new utilty 'ncpsend' which allow sending broadcast messages to NetWare users. And, as always, few minor bug fixes. BTW, Eric Schmidt (Novell's CEO), give a promise for help in NDS support implementation. Initial NDS support will be included in ver 1.4. URL: ftp://ftp.butya.kz/pub/nwlib/ncplib.tar.gz P.S. Does someone runs ncplib on 2 processors machine with SMP kernel ? -- Boris Popov http://www.butya.kz/~bp/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 4:37:36 1999 Delivered-To: freebsd-net@freebsd.org Received: from kamna.i.cz (kamna.i.cz [193.85.255.30]) by hub.freebsd.org (Postfix) with SMTP id A467D14EA4 for ; Sat, 27 Mar 1999 04:37:33 -0800 (PST) (envelope-from mm@i.cz) Received: (qmail 21487 invoked from network); 27 Mar 1999 12:37:12 -0000 Received: from woody.i.cz (@193.85.255.60) by kamna.i.cz with SMTP; 27 Mar 1999 12:37:12 -0000 Content-Length: 1832 Message-ID: X-Mailer: XFMail 1.3 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-2 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199903262137.PAA06872@carp.gbr.epa.gov> Date: Sat, 27 Mar 1999 13:37:12 +0100 (MET) Reply-To: mm@i.cz From: Martin Machacek To: freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 26-Mar-99 Mike Jenkins wrote: > On Fri, 26 Mar 1999 Martin Machacek wrote: >> Layer 4 switch is a pure marketing bullshit. > > If I understand layer 4 switches correctly, they switch > at the tcp/udp port number layer. I could therefore slip > a layer 4 switch between my router and my lan, and program > it to redirect all incoming 25/tcp smtp connections to a > mail filter host. I find that rather useful. I'm sure > some folks use them for 80/tcp http redirection for web > caching. Aren't these useful applicatons? Sure, but they can be quite easily achieved with "convetional" router (or a good layer 3 switch). The router could be also a PC with decent Unix like for example FreeBSD. What these applications require is capability to do routing (switching) decisions based on other information than destination IP address. This feature is usually called policy routing and you can find it in most routers from major vendors. So, I think that so called layer 4 switches are just IP routers with policy routing capability. I doubt that layer 4 switching is being implemented in hardware. > I realize routers can be programmed to do this but who wants > to load down (or misconfigure) the router for this chore. Policy routing is not a misconfiguration. For sure it can put significant load on the router. Who wants to buy another box ... > A dual-homed unix box such as FreeBSD can also do this using > redirection in packet filtering but that usually requires > splitting the network into 2 IP networks (yes i've heard > of dummynet/bridge but that is work in progress). I think > a network appliance like a layer 4 switch would be the right > tool for the job. Why you think you have to split your network and actually what's wrong with splitting IP networks anyway? Martin --- [PGP KeyID F3F409C4]] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 7: 9:12 1999 Delivered-To: freebsd-net@freebsd.org Received: from obie.softweyr.com (unknown [204.68.178.33]) by hub.freebsd.org (Postfix) with ESMTP id 46AA214C8E for ; Sat, 27 Mar 1999 07:09:09 -0800 (PST) (envelope-from wes@softweyr.com) Received: from softweyr.com (wes@zaphod.softweyr.com [204.68.178.35]) by obie.softweyr.com (8.8.8/8.8.8) with ESMTP id IAA01510; Sat, 27 Mar 1999 08:08:46 -0700 (MST) (envelope-from wes@softweyr.com) Message-ID: <36FCF47D.403DECF3@softweyr.com> Date: Sat, 27 Mar 1999 08:08:45 -0700 From: Wes Peters Organization: Softweyr llc X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Jonathan Lemon Cc: freebsd-net@freebsd.org Subject: Re: switch vs bridge (fwd) References: <199903270103.TAA14685@free.pcs> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Jonathan Lemon wrote: > > In article you write: > >Mike Jenkins wrote: > >> I find that rather useful. I'm sure > >> some folks use them for 80/tcp http redirection for web > >> caching. > > > >Well, more likely for bandwidth/performance management and hot failover. > > That's actually what the particular switch I have to deal with is > doing; it an Alteon switch, set up as a transparent web proxy server. > As I understand it, it intercepts all web traffic from the campus > and directs it to a pool of proxies. I suppose it's useful for what > it does. > > How would a layer-3 switch do the same thing? VRRP. Gee, why do something with a standard when we can "extend and extinguish?" -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC http://www.softweyr.com/~softweyr wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 7:14:17 1999 Delivered-To: freebsd-net@freebsd.org Received: from obie.softweyr.com (unknown [204.68.178.33]) by hub.freebsd.org (Postfix) with ESMTP id 610D714F4F for ; Sat, 27 Mar 1999 07:14:14 -0800 (PST) (envelope-from wes@softweyr.com) Received: from softweyr.com (wes@zaphod.softweyr.com [204.68.178.35]) by obie.softweyr.com (8.8.8/8.8.8) with ESMTP id IAA01522; Sat, 27 Mar 1999 08:13:38 -0700 (MST) (envelope-from wes@softweyr.com) Message-ID: <36FCF5A2.8F3F11E9@softweyr.com> Date: Sat, 27 Mar 1999 08:13:38 -0700 From: Wes Peters Organization: Softweyr llc X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: Mike Jenkins Cc: freebsd-net@freebsd.org, mm@i.cz Subject: Re: switch vs bridge (fwd) References: <199903270149.TAA07234@carp.gbr.epa.gov> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Mike Jenkins wrote: > > Wes Peters wrote: > > The you either don't understand the job, or don't understand the (very > > limited) capabilities of these so-called layer 4 switches. It's not that > > it's a bad idea, just that there are a couple of vendors out there giving > > the idea a bad name with their implementations. > > I think the concept of switching based on udp/tcp port is useful. > One thing I'm not clear on is at what layer is the switch gathering > the packets? Is it at the data link (Ethernet) layer or the network > layer (IP)? I was assuming the data link layer to allow transparency. > If it is the network layer then the switch is your router as Wes said. > Do layer 3 switches have the ability to redirect packets based on > udp/tcp port numbers? That is, by definition, layer 4. Layer 3 is the IP address layer, and no "port" information occurs there. As I said above, it's not that layer 4 switching is a bad idea, it's just that the current implementation from some of these vendors is a bit weak, and none are based on standard protocols. And it *is* a bad idea to have every switch on your network peeking all the way into the layer 4 information of EVERY packet flowing through, that means you've bought a couple thousand times more packet processing power than you really need. The best way to do it would be to stick a layer-4 balancing engine inside a switch chassis, managing ports via the switch backplane. AFAIK nobody has done this yet, but it's such a darned good idea, I know somebody who might. ;^) -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC http://www.softweyr.com/~softweyr wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 7:21:26 1999 Delivered-To: freebsd-net@freebsd.org Received: from obie.softweyr.com (unknown [204.68.178.33]) by hub.freebsd.org (Postfix) with ESMTP id 6DEBF14F6B for ; Sat, 27 Mar 1999 07:21:19 -0800 (PST) (envelope-from wes@softweyr.com) Received: from softweyr.com (wes@zaphod.softweyr.com [204.68.178.35]) by obie.softweyr.com (8.8.8/8.8.8) with ESMTP id IAA01548; Sat, 27 Mar 1999 08:20:36 -0700 (MST) (envelope-from wes@softweyr.com) Message-ID: <36FCF743.F6530F5C@softweyr.com> Date: Sat, 27 Mar 1999 08:20:35 -0700 From: Wes Peters Organization: Softweyr llc X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.1-RELEASE i386) X-Accept-Language: en MIME-Version: 1.0 To: mm@i.cz Cc: freebsd-net@FreeBSD.ORG Subject: Re: switch vs bridge (fwd) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Martin Machacek wrote: > > On 26-Mar-99 Mike Jenkins wrote: > > On Fri, 26 Mar 1999 Martin Machacek wrote: > >> Layer 4 switch is a pure marketing bullshit. > > > > If I understand layer 4 switches correctly, they switch > > at the tcp/udp port number layer. I could therefore slip > > a layer 4 switch between my router and my lan, and program > > it to redirect all incoming 25/tcp smtp connections to a > > mail filter host. I find that rather useful. I'm sure > > some folks use them for 80/tcp http redirection for web > > caching. Aren't these useful applicatons? > > Sure, but they can be quite easily achieved with "convetional" router (or a > good layer 3 switch). The router could be also a PC with decent Unix like for > example FreeBSD. What these applications require is capability to do > routing (switching) decisions based on other information than destination IP > address. This feature is usually called policy routing and you can find it in > most routers from major vendors. So, I think that so called layer 4 switches > are just IP routers with policy routing capability. I doubt that layer 4 > switching is being implemented in hardware. Not yet. It soon will be. You're not going to believe the speed. > > I realize routers can be programmed to do this but who wants > > to load down (or misconfigure) the router for this chore. > > Policy routing is not a misconfiguration. For sure it can put significant > load on the router. Who wants to buy another box ... > > > A dual-homed unix box such as FreeBSD can also do this using > > redirection in packet filtering but that usually requires > > splitting the network into 2 IP networks (yes i've heard > > of dummynet/bridge but that is work in progress). I think > > a network appliance like a layer 4 switch would be the right > > tool for the job. The realy problem with a layer 4 switch is that it doesn't have any knowlege of the data stream it's handling, it just doles out packets based on some really simple rules. A load balancer that has some understand of the underlying data is probably a much better way of doing things. In the case of HTTP, a load balancer can keep a client connection on the same server, in case the server is maintaining some connection information, and can make decisions on which server is the least busy when a new connection comes in, based *at least* on which is handling the fewest number of packets at the moment. Layer 4 switches don't do either of these, they just dole out packets in (typically) round-robin fashion. > Why you think you have to split your network and actually what's wrong with > splitting IP networks anyway? Nothing is wrong with splitting IP networks, you just don't need to complicate it with a bunch of goofy routing configurations. If you want to keep the engineering staff from sniffing packets over in HR, stick everyone on switch ports and manage your VLANs carefully. ;^) -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC http://www.softweyr.com/~softweyr wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 7:42:50 1999 Delivered-To: freebsd-net@freebsd.org Received: from sumatra.americantv.com (sumatra.americantv.com [207.170.17.37]) by hub.freebsd.org (Postfix) with ESMTP id 1CAAB150A6 for ; Sat, 27 Mar 1999 07:42:47 -0800 (PST) (envelope-from jlemon@americantv.com) Received: from right.PCS (right.PCS [148.105.10.31]) by sumatra.americantv.com (8.8.5/8.8.5) with ESMTP id JAA28108; Sat, 27 Mar 1999 09:42:27 -0600 (CST) Received: (from jlemon@localhost) by right.PCS (8.6.13/8.6.4) id JAA07119; Sat, 27 Mar 1999 09:41:56 -0600 Message-ID: <19990327094156.30720@right.PCS> Date: Sat, 27 Mar 1999 09:41:56 -0600 From: Jonathan Lemon To: Wes Peters Cc: freebsd-net@freebsd.org Subject: Re: switch vs bridge (fwd) References: <199903270103.TAA14685@free.pcs> <36FCF47D.403DECF3@softweyr.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.61.1 In-Reply-To: <36FCF47D.403DECF3@softweyr.com>; from Wes Peters on Mar 03, 1999 at 08:08:45AM -0700 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mar 03, 1999 at 08:08:45AM -0700, Wes Peters wrote: > Jonathan Lemon wrote: > > > > In article you write: > > >Mike Jenkins wrote: > > >> I find that rather useful. I'm sure > > >> some folks use them for 80/tcp http redirection for web > > >> caching. > > > > > >Well, more likely for bandwidth/performance management and hot failover. > > > > That's actually what the particular switch I have to deal with is > > doing; it an Alteon switch, set up as a transparent web proxy server. > > As I understand it, it intercepts all web traffic from the campus > > and directs it to a pool of proxies. I suppose it's useful for what > > it does. > > > > How would a layer-3 switch do the same thing? > > VRRP. Gee, why do something with a standard when we can "extend and > extinguish?" "The wonderful thing about standards is that there are so many to choose from". Isn't VRRP RFC-2338? In this case, what I'm looking for is not to specify a first hop router, but some way to effectively insert a caching proxy for all outbound web traffic. -- Jonathan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 11:40:31 1999 Delivered-To: freebsd-net@freebsd.org Received: from smtp05.wxs.nl (smtp05.wxs.nl [195.121.6.57]) by hub.freebsd.org (Postfix) with ESMTP id A637A14C3E for ; Sat, 27 Mar 1999 11:40:28 -0800 (PST) (envelope-from asmodai@wxs.nl) Received: from daemon.ninth-circle.org ([195.121.57.41]) by smtp05.wxs.nl (Netscape Messaging Server 3.61) with ESMTP id AAA3539; Sat, 27 Mar 1999 20:40:07 +0100 Received: from daemon.ninth-circle.org (abaddon@daemon [192.168.0.1]) by daemon.ninth-circle.org (8.9.3/8.9.3) with ESMTP id UAA82576; Sat, 27 Mar 1999 20:40:10 +0100 (CET) (envelope-from asmodai@wxs.nl) Message-ID: X-Mailer: XFMail 1.3 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <4.1.19990326233717.009c8210@mail.dnai.com> Date: Sat, 27 Mar 1999 20:40:09 +0100 (CET) Organization: Ninth Circle Enterprises From: Jeroen Ruigrok/Asmodai To: Mike Thompson Subject: RE: FreeBSD as a router Cc: freebsd-net@freebsd.org Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 27-Mar-99 Mike Thompson wrote: > This is probably a stupid question (not that it ever stopped > me before), but is a FreeBSD system capable of running as a > high-capacity Internet router? I would like to run a number > of FreeBSD servers running a web application behind a FreeBSD > system acting as a router/firewall. Any examples of web sites > doing this would be great. We do so for a 2 Mbit line at work, the line gets the data in through a Cisco and then gets through 100 Mbit to the FreeBSD firewall/router. Had an uptime of avg 300 days before we had to shut it down due to UPS relocation. Otherwise it reacted very well... HTH, --- Jeroen Ruigrok van der Werven asmodai(at)wxs.nl The idea does not replace the work... Network/Security Specialist *BSD: Powered by Knowledge & Know-how To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 12:18:44 1999 Delivered-To: freebsd-net@freebsd.org Received: from freja.webgiro.com (10.0.29.209.212.in-addr.arpa [212.209.29.10]) by hub.freebsd.org (Postfix) with ESMTP id 6FB2814A23; Sat, 27 Mar 1999 12:18:35 -0800 (PST) (envelope-from abial@webgiro.com) Received: by freja.webgiro.com (Postfix, from userid 1001) id 3C50C18C6; Sat, 27 Mar 1999 21:18:16 +0100 (CET) Received: from localhost (localhost [127.0.0.1]) by freja.webgiro.com (Postfix) with ESMTP id 392AD4999; Sat, 27 Mar 1999 21:18:16 +0100 (CET) Date: Sat, 27 Mar 1999 21:18:06 +0100 (CET) From: Andrzej Bialecki To: Greg Lehey Cc: Jesse , "Daniel C. Sobral" , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: Taking panic dumps (was: 3.1-STABLE dies on 40+ connects (resolved)) In-Reply-To: <19990327174738.B425@lemis.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 Sat, 27 Mar 1999, Greg Lehey wrote: > On Thursday, 25 March 1999 at 16:09:44 -0800, Jesse wrote: > > > >>> Thanks everybody for your help! FreeBSD is great and it's 80% about the > >>> community. Not that I didn't always know it. =) > >> > >> Notice that a core dump + kernel with symbols would still be very > >> useful to find *where* in our code is the bug. > > > > I asked in other messages if anyone still wanted me to do that. No one > > replied. Anyway, I'll do this at an off hour and get a core dump. > > OK. I really should put this in the FAQ, since it's really important. I agree - it's really important, that's why it's in the handbook. :-) Andrzej Bialecki // WebGiro AB, Sweden (http://www.webgiro.com) // ------------------------------------------------------------------- // ------ FreeBSD: The Power to Serve. http://www.freebsd.org -------- // --- Small & Embedded FreeBSD: http://www.freebsd.org/~picobsd/ ---- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 13:17:28 1999 Delivered-To: freebsd-net@freebsd.org Received: from i.caniserv.com (i.caniserv.com [139.142.95.1]) by hub.freebsd.org (Postfix) with SMTP id 1A19B15248 for ; Sat, 27 Mar 1999 13:16:43 -0800 (PST) (envelope-from Darcy@ok-connect.com) Received: (qmail 11485 invoked from network); 27 Mar 1999 21:16:23 -0000 Received: from ccliii.caniserv.com (HELO dbitech) (darcyb@139.142.95.253) by 139.142.95.10 with SMTP; 27 Mar 1999 21:16:23 -0000 Message-Id: <3.0.32.19990327131741.01d08030@mail.ok-connect.com> X-Sender: darcyb@mail.ok-connect.com X-Mailer: Windows Eudora Pro Version 3.0 (32) Date: Sat, 27 Mar 1999 13:17:42 -0800 To: freebsd-net@freebsd.org From: Darcy Buskermolen Subject: REFreeBSD as a router Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I run a FreeBSD router/firewall on a fractional T3 connection to the internet, the system is a P100 with 32MB ram and 6 NIC's and is also using Dummynet to do some minor pipes, uptimes were in excess of 200 days until we moved offices. I would recommend FreeBSD in ANY hi volume application. >This is probably a stupid question (not that it ever stopped >me before), but is a FreeBSD system capable of running as a >high-capacity Internet router? I would like to run a number >of FreeBSD servers running a web application behind a FreeBSD >system acting as a router/firewall. Any examples of web sites >doing this would be great. > >Mike Thompson > > > > >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 Sat Mar 27 13:31:51 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 9693114F15 for ; Sat, 27 Mar 1999 13:31:46 -0800 (PST) (envelope-from mike@sentex.net) Received: from ospf-wat.sentex.net (ospf-wat.sentex.net [209.167.248.81]) by granite.sentex.net (8.8.8/8.6.9) with SMTP id QAA16848; Sat, 27 Mar 1999 16:31:22 -0500 (EST) From: mike@sentex.net (Mike Tancsa) To: miket@dnai.com (Mike Thompson) Cc: freebsd-net@FreeBSD.ORG Subject: Re: FreeBSD as a router Date: Sat, 27 Mar 1999 21:40:24 GMT Message-ID: <36fd12fb.3761327633@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 27 Mar 1999 02:44:45 -0500, in sentex.lists.freebsd.misc you wrote: >This is probably a stupid question (not that it ever stopped >me before), but is a FreeBSD system capable of running as a >high-capacity Internet router? I would like to run a number >of FreeBSD servers running a web application behind a FreeBSD >system acting as a router/firewall. Any examples of web sites >doing this would be great. Have a look through the various archives. You will see this topic discussed several times. Also, you will need to be a little more specific about the term 'high-capacity'.. In an industry riddled with high-hyperbole, what does high capacity mean ? 10Mb ? 100Mb ? 155Mb ? 1 Gigabit ? ftp.cdrom.com is probably a good example, as is www.yahoo.com which also makes use of FreeBSD. As an ethernet router, I can push through over 10Mbits comfortably through my router with a dozen or so firewall rules as well as running gated with 2 views. If you are talking about anything faster, you are best off to test it yourself to see if it will meet your needs or not. ---Mike Mike Tancsa (mdtancsa@sentex.net) Sentex Communications Corp, Waterloo, Ontario, Canada To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 14: 0:30 1999 Delivered-To: freebsd-net@freebsd.org Received: from inner.net (avarice.inner.net [199.33.248.2]) by hub.freebsd.org (Postfix) with ESMTP id D030914F0C for ; Sat, 27 Mar 1999 14:00:28 -0800 (PST) (envelope-from cmetz@inner.net) Received: from inner.net (cmetz.cstone.net [205.197.102.217]) by inner.net (8.9.1/8.9.1) with ESMTP id VAA08726; Sat, 27 Mar 1999 21:56:06 GMT Message-Id: <199903272156.VAA08726@inner.net> To: mike@sentex.net (Mike Tancsa) Cc: freebsd-net@FreeBSD.ORG Subject: Re: FreeBSD as a router In-reply-to: Your message of "Sat, 27 Mar 1999 21:40:24 GMT." <36fd12fb.3761327633@mail.sentex.net> X-Copyright: Copyright 1999, Craig Metz, All Rights Reserved. X-Reposting: With explicit permission only Date: Sat, 27 Mar 1999 16:59:08 -0500 From: Craig Metz Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In message <36fd12fb.3761327633@mail.sentex.net>, you write: >On 27 Mar 1999 02:44:45 -0500, in sentex.lists.freebsd.misc you wrote: > >>This is probably a stupid question (not that it ever stopped >>me before), but is a FreeBSD system capable of running as a >>high-capacity Internet router? I would like to run a number >>of FreeBSD servers running a web application behind a FreeBSD >>system acting as a router/firewall. Any examples of web sites >>doing this would be great. > >Have a look through the various archives. You will see this topic >discussed several times. Also, you will need to be a little more specific >about the term 'high-capacity'.. In an industry riddled with >high-hyperbole, what does high capacity mean ? >10Mb ? 100Mb ? 155Mb ? 1 Gigabit ? ftp.cdrom.com is probably a good >example, as is www.yahoo.com which also makes use of FreeBSD. As an >ethernet router, I can push through over 10Mbits comfortably through my >router with a dozen or so firewall rules as well as running gated with 2 >views. If you are talking about anything faster, you are best off to test >it yourself to see if it will meet your needs or not. What really matters here is the application. If the original poster was thinking about using FreeBSD-based commidity PCs for the core routers of a large ISP, well, he can keep thinking that (just don't do it!). The hardware can't move the bits fast enough and keep the interfaces fed, FreeBSD isn't able to forward packets that fast, and standard GateD can't handle that sort of routing load/complexity. Several "real" router vendors (e.g., Torrent, Juniper) use very special hardware (including hardware fast-path forwarding) and special routing software with FreeBSD in the middle, so the problem is not FreeBSD itself but the pieces above and below. If the original poster was thinking about using FreeBSD-based commodity PCs for small-to-midrange routers in an environment where cost/flexibility is important and it doesn't have to deliver every last bit of line capacity, then it might be a great choice. Research networks (e.g., vBNS and CAIRN) are deploying FreeBSD-based PCs as midrange WAN routers (OC3 ATM on one side, 100Mb/s Ethernet on the other), and in the process, are trying to push what can be done on such a platform. A FreeBSD (*BSD or Linux, too) PC with a quad-Tulip board can build you a decent four-/eight-/twelve-port 100Mb/s router for a lot less cash than a "real" router vendor would charge you, which might be reasonable for a mid-sized business with a lot of subnets and not so much traffic. This might fit some people's definition of high capacity. In both scenarios, a FreeBSD-based PC might not actually be able to move the bits fast enough to run all four ports at capacity (PC NIC boards tend not to have useful amounts of buffer on-board, and the PCI bus can only service one device at a time), but that might not be really necessary, and the difference in cost might really be worth it. And with FreeBSD, you have the source code. -Craig To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 18:15:50 1999 Delivered-To: freebsd-net@freebsd.org Received: from allegro.lemis.com (allegro.lemis.com [192.109.197.134]) by hub.freebsd.org (Postfix) with ESMTP id D9F13150BD; Sat, 27 Mar 1999 18:14:34 -0800 (PST) (envelope-from grog@freebie.lemis.com) Received: from freebie.lemis.com (freebie.lemis.com [192.109.197.137]) by allegro.lemis.com (8.9.1/8.9.0) with ESMTP id LAA17065; Sun, 28 Mar 1999 11:44:14 +0930 (CST) Received: (from grog@localhost) by freebie.lemis.com (8.9.3/8.9.0) id LAA70046; Sun, 28 Mar 1999 11:44:11 +0930 (CST) Message-ID: <19990328114410.U53452@lemis.com> Date: Sun, 28 Mar 1999 11:44:10 +0930 From: Greg Lehey To: Andrzej Bialecki Cc: Jesse , "Daniel C. Sobral" , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Debug kernel by default? (was: Taking panic dumps (was: 3.1-STABLE dies on 40+ connects (resolved))) References: <19990327174738.B425@lemis.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.2i In-Reply-To: ; from Andrzej Bialecki on Sat, Mar 27, 1999 at 09:18:06PM +0100 WWW-Home-Page: http://www.lemis.com/~grog Organization: LEMIS, PO Box 460, Echunga SA 5153, Australia Phone: +61-8-8388-8286 Fax: +61-8-8388-8725 Mobile: +61-41-739-7062 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Saturday, 27 March 1999 at 21:18:06 +0100, Andrzej Bialecki wrote: > On Sat, 27 Mar 1999, Greg Lehey wrote: > >> On Thursday, 25 March 1999 at 16:09:44 -0800, Jesse wrote: >>> >>>>> Thanks everybody for your help! FreeBSD is great and it's 80% about the >>>>> community. Not that I didn't always know it. =) >>>> >>>> Notice that a core dump + kernel with symbols would still be very >>>> useful to find *where* in our code is the bug. >>> >>> I asked in other messages if anyone still wanted me to do that. No one >>> replied. Anyway, I'll do this at an off hour and get a core dump. >> >> OK. I really should put this in the FAQ, since it's really important. > > I agree - it's really important, that's why it's in the handbook. :-) Well, there's a threshold difference between the handbook and the FAQ. In addition, the description in the handbook is rather confusing. I intend to overhaul it once the handbook comes out of its freeze (hopefully soon). In that connection, any comments about changing the default way of building a kernel to create a debug kernel and a stripped copy, and install the stripped copy? It would require about 10 MB more storage and a little more time to build the kernel, but since kgdb is useless without the debug symbols, and disk space is cheap, it seems to me that it would be worthwhile. Greg -- See complete headers for address, home page and phone numbers finger grog@lemis.com for PGP public key To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 19:10: 0 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail.rwth-aachen.de (mail.RWTH-Aachen.DE [137.226.144.9]) by hub.freebsd.org (Postfix) with ESMTP id BCD5D1522C for ; Sat, 27 Mar 1999 19:09:30 -0800 (PST) (envelope-from heinig@hdz-ima.rwth-aachen.de) Received: from HDZ-IMA.RWTH-Aachen.de (majestix.hdz-ima.RWTH-Aachen.DE) by mail.rwth-aachen.de (PMDF V5.1-12 #30440) with ESMTP id <01J9CMBTB7CQ0003GK@mail.rwth-aachen.de> for net@FreeBSD.ORG; Sun, 28 Mar 1999 04:10:09 +0100 Received: from MAJESTIX/MAIL by HDZ-IMA.RWTH-Aachen.de (Mercury 1.20); Sun, 28 Mar 1999 04:11:36 +0000 Received: from MAIL by MAJESTIX (Mercury 1.20); Sun, 28 Mar 1999 04:11:23 +0000 Received: from hdz-ima.rwth-aachen.de by HDZ-IMA.RWTH-Aachen.de (Mercury 1.20) with ESMTP; Sun, 28 Mar 1999 04:11:15 +0000 Date: Sun, 28 Mar 1999 05:09:27 +0200 From: Gerald Heinig Subject: Re: OSI layering Query.. Please Help ME To: danny@alpha.net.au Cc: net@FreeBSD.ORG, Tony Finch Message-id: <36FD9D67.4BFE39D@hdz-ima.rwth-aachen.de> Organization: Informatik im Maschinenbau / Hochschuldidaktisches Zentrum, RWTH Aachen MIME-version: 1.0 X-Mailer: Mozilla 4.5 [en] (X11; I; FreeBSD 2.2.7-RELEASE i386) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <199903262358.JAA27392@sydney.alpha.net.au> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Danny Ho wrote: > > Hi everyone I am sort of new( 1 year experience in this ISP job) > > Anyway, I have one query about OSI layering. > > What are some of the advantages of OSI layer as seen in the OSI model?? Advantages? Hmmm. I suppose perhaps somewhat finer granularity than the IP model. It also defines layers above 4 (transport layer) which AFAIK IP doesn't. It's perhaps somewhat more complete. I certainly make no claims to any great expertise on the subject. Comments, anyone? > > And what exactly are PDU? Protocol Data Units. The information packages/packets that get sent up/down the protocol stacks during protocol operation/data transmission/reception. For example, a TPDU (transport protocol data unit) contains the TSAP address (transport service access point address - corresponds loosely to the "port number" in IP) and the "payload" ie. data you're trying to transmit. This gets sent down (if you're transmitting) to the NSAP (network service access point - equivalent to a network interface in IP) which adds its NSAP address (which corresponds to the IP address in IP). This then gets sent down to the link layer (layer 2) to the LSAP (link layer service access point, ie. the network card) which adds its LSAP address (ie. MAC/hardware/ethernet address) plus header/trailer/FCS (frame check sequence) and sends the caboodle off onto the wire. > > What exactly is SAP?? Service Access Point. The place where you can access a service. A telephone socket is the SAP for a connection-oriented voice data transfer service. IP mail exchange has an (IP) transport service access point address of 25 ie.it uses IP port 25 for mail exchange. Your machine's network interface would be the NSAP, the interface's IP address would correspond to its NSAP address. Note: *CORRESPOND*!!! There is such a thing as an NSAP address and it is *NOT* the same as an IP address!!! we're talking analogies here! > > I am really confused help me. I hope the above is correct. It's certainly what I was taught, and what it says in the book about networking by Fred Halsall "Data Communications, computer networks and OSI". I believe there's a second edition out (I've got the first one, and it's somewhat dated...). I admit freely that I'm a bit shaky on the subject: the book gives no concrete examples and is pretty abstract for the most part. Corrections/amendments are *most* appreciated!! Anyway, if it's right (I'm fairly sure it is!), I hope that helps Cheers, Gerald -- "Would you like to buy an encyclopaedia to help your child get to college?" "He doesn't need it. He takes the bus!" To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 20:51:55 1999 Delivered-To: freebsd-net@freebsd.org Received: from poboxer.pobox.com (unknown [208.149.16.9]) by hub.freebsd.org (Postfix) with ESMTP id 1C3AF14C42 for ; Sat, 27 Mar 1999 20:51:51 -0800 (PST) (envelope-from alk@poboxer.pobox.com) Received: (from alk@localhost) by poboxer.pobox.com (8.9.3/8.9.1) id WAA54371; Sat, 27 Mar 1999 22:51:13 -0600 (CST) (envelope-from alk) From: Anthony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sat, 27 Mar 1999 22:51:12 -0600 (CST) X-Face: \h9Jg:Cuivl4S*UP-)gO.6O=T]]@ncM*tn4zG);)lk#4|lqEx=*talx?.Gk,dMQU2)ptPC17cpBzm(l'M|H8BUF1&]dDCxZ.c~Wy6-j,^V1E(NtX$FpkkdnJixsJHE95JlhO 5\M3jh'YiO7KPCn0~W`Ro44_TB@&JuuqRqgPL'0/{):7rU-%.*@/>q?1&Ed Reply-To: alk@pobox.com To: miket@dnai.com Cc: freebsd-net@FreeBSD.ORG Subject: PPP as a bridge over the Internet References: <4.1.19990326173538.00a46490@mail.dnai.com> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <14077.43226.915430.396131@avalon.east> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Quoth Mike Thompson on Fri, 26 March: : I've been considering using FreeBSD 2.2.8 with PPP OVER TCP to : bridge two remote private networks over the Internet. The following : caveat appears in the PPP man page. Does anyone have real-world : experience with this being a problem? I've never used it at speeds over 56k. The encapsulation seems to add about 40ms to the ping latency, using iijppp on one end and kernel ppp on the other, but I've never observed a significant performance degradation due to packet loss. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message From owner-freebsd-net Sat Mar 27 20:57:41 1999 Delivered-To: freebsd-net@freebsd.org Received: from mail-out2.apple.com (mail-out2.apple.com [17.254.0.51]) by hub.freebsd.org (Postfix) with ESMTP id D592F14C20 for ; Sat, 27 Mar 1999 20:57:39 -0800 (PST) (envelope-from justin@walker3.apple.com) Received: from mailgate1.apple.com (A17-128-100-225.apple.com [17.128.100.225]) by mail-out2.apple.com (8.8.5/8.8.5) with ESMTP id UAB25030 for ; Sat, 27 Mar 1999 20:54:40 -0800 Received: from scv2.apple.com (scv2.apple.com) by mailgate1.apple.com (mailgate1.apple.com- SMTPRS 2.0.15) with ESMTP id ; Sat, 27 Mar 1999 20:54:38 -0800 Received: from walker3.apple.com (walker3.apple.com [17.219.24.201]) by scv2.apple.com (8.9.3/8.9.3) with ESMTP id UAA19088; Sat, 27 Mar 1999 20:54:36 -0800 Received: by walker3.apple.com (8.9.1/8.9.1) id UAA00698; Sat, 27 Mar 1999 20:54:39 -0800 (PST) Message-Id: <199903280454.UAA00698@walker3.apple.com> To: danny@alpha.net.au Subject: Re: OSI layering Query.. Please Help ME Cc: net@FreeBSD.ORG Date: Sat, 27 Mar 1999 20:54:38 -0800 From: "Justin C. Walker" Reply-To: justin@apple.com X-Mailer: by Apple MailViewer (2.105.dev) Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > From: Danny Ho > Date: 1999-03-26 15:47:31 -0800 > To: net@FreeBSD.ORG, Tony Finch > Subject: OSI layering Query.. Please Help ME > X-MSMail-Priority: Normal > X-Priority: 3 > Delivered-to: freebsd-net@freebsd.org > X-Mailer: Microsoft Internet Mail 4.70.1155 > X-Loop: FreeBSD.org > > Hi everyone I am sort of new( 1 year experience in this ISP job) > > Anyway, I have one query about OSI layering. > > What are some of the advantages of OSI layer as seen in the OSI model?? Perhaps the biggest advantage of the OSI model is the language it brings to the design of protocols. Opinions will vary as to the advantages in actual implementation, but generally, some adherence to the layering model makes your code cleaner and more maintainable. You need to weigh that aspect against that of performance, which is where things will get interesting. You can find detailed discussions of the model, and its applicabiliity to IP, in many places. See, e.g., Stallings, "Handbook of Computer Communications Standards, V. 1, The Open Systems Interconnect model ..." (MacMillan/Stallings); Stevens' "TCP/IP Illustrated, V1" (Addison Wesley), discusses layering as it applies specifically to TCP and IP (i.e., up through what OSI calls the transport layer). > And what exactly are PDU? Protocol Data Units - this is OSI-speak for "packets", although it allows you to talk about packets at the various protocol layers (e.g., an IP packet as distinct from a TCP packet or an ethernet packet). > What exactly is SAP?? Service Access Point - it's a protocol addressing term. E.g., in IP terms, an SAP at the network layer would be a port (and is known as an NSAP). You'd need to delve more deeply into the OSI model to appreciate the subtlety of the concept :-}. > I am really confused help me. Hope this helps. Regards, Justin Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | Manager, CoreOS Networking | Men are from Earth. Apple Computer, Inc. | Women are from Earth. 2 Infinite Loop | Deal with it. Cupertino, CA 95014 | *-------------------------------------*-------------------------------* To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message