About The Linux Modular Bridge And STP Uwe Böhme

Johann-Heinrich-Abt-Straße 7 95213 Münchberg Germany +49/9251 960877 +49/9251 960878 uwe@bnhof.de

Lennert Buytenhenk He made the bridge fly. bridge code maintainer and developer gnu.org

buytenh@gnu.org

The source, a lot of useful advice,documentation of STP, comments to message logs, typo correction,devices and encouragement. 2000 Uwe Böhme Release v0.04 v0.04 11 January 2001 U.B. Changed Lennert`s Bridge Homepage URL; added NIC to list. v0.03 17 July 2000 U.B. Overwork pdf. Download links in doc. v0.02 16 July 2000 U.B. Fixed broken graphics in html dsl. Prepared pdf. Typos. v0.01 25 June 2000 U.B. Changes name from BRIDGE-HOWTO to &tit; (avoid interference with BRIDGE-HOWTO by Christopher Cole) and kill version 1.xx. Lennert Buytenhenk announced as coauthor. v0.00 01 June 2000 U.B. Initial Release. All files are copyrighted by their mentioned author or organization as mentioned in the file. This file may be distributed or modified according to the terms of LDP Manifesto. is based on the introduction to the BRIDGE-HOWTO by Christopher Cole

cole@coledd.com

Published v1.11, 7 September 1998 with LDP-Copyright. This document describes how to setup a bridge with the recent kernel patches and brctl utility by Lennert Buytenhek. and tries to explain about the STP implementation in this code. With developer kernel 2.3.47 the new bridging code is part of the mainstream. There are patches for stable kernels 2.2.14 to 2.2.16, where each is also available as a ipchains-patch. Copyright (c) 2000 by Uwe Böhme. This document may be distributed only subject to the terms and conditions set forth in the LDP License available at &ldp; Official url is &hbr;. With developer kernel 2.3.47 the new bridging code is part of the mainstream. The Bridge-Mailinglist is homed at http://www.math.leidenuniv.nl/mailman/listinfo/bridge. This document has it's official homepage at &udp;&dir;/&tit;/. It's a part of the Linux Documentation Project located at &ldp;. Build environment as tar.gziped file &udp;&dir;/&tit;.tar.gz HTML-gziped file &udp;&dir;/&tit;.html.tar.gz PDF-gziped file &udp;&dir;/&tit;.pdf.gz PS-gziped file &udp;&dir;/&tit;.ps.gz A bridge is a device that separates two or more network segments within one logical network (e.g. a single IP-subnet). A bridge is usually placed between two separate groups of computers that talk with each other, but not that much with the computers in the other group. A good example of this is to consider a cluster of Macintoshes and a cluster of Unix machines. Both of these groups of machines tend to be quite chatty amongst themselves, and the traffic they produce on the network causes collisions for the other machines who are trying to speak to one another. The job of the bridge is to examine the destination of the data packets one at a time and decide whether or not to pass the packets to the other side of the Ethernet segment. The result is a faster, quieter network with less collisions. The bridging code decides whether to bridge data or to drop it not by looking at the protocol type (IP, IPX, NetBEUI), but by looking at the &mac;-address unique to each NIC. It's vital to understand that a bridge is neither a router nor a fire-wall. Spoken in simple term a bridge behaves like a network switch (i.e. Layer 2 Switch), making it a transparent network component (which is not absolutely true, but nearly). Read more about this at . In addition, you can overcome hardware incompatibilities with a bridge, without leaving the address-range of your IP-net or subnet. E.g. it's possible to bridge between different physical media like 10 Base T and 100 Base TX. My personal reason for starting to set up a bridge was that in my work I had to connect Fast Ethernet components to a existing HP Voice Grade network, which is a proprietary networking standard. STP The Spanning Tree Protocol is a nifty method of keeping Ethernet devices connected in multiple paths working. The participating switches negotiate the shortest available path by STP. This feature will be discussed in . Multiple Bridge Instances Multiple bridge instances allow you to have more than one bridge on your box up and running, and to control each instance separately. Fire-walling There is a patch to the bridging code which allows you to use IP chains on the interface inside a bridge. More info about this you'll find at . There is a number of rules you are not allowed to break (otherwise your bridge will do). A port can only be a member of one bridge. A bridge knows nothing about routes. A bridge knows nothing about higher protocols than &arp;. That's the reason why it can bridge any possible protocol possibly running on your Ethernet. No matter how many ports you have in your logical bridge, it's covered by only one logical interface As soon as a port (e.g. a NIC) is added to a bridge you have no more direct control about it. If one of the points mentioned above is not clear to you now, don't continue reading. Read the documents listed in first. If you ever tried to ping an unmanaged switch, you will know that it doesn't work, because you don't have a IP-address for it. To switch datagrams it doesn't need one. The other thing is if you want to manage the switch. It's too much strain, to take a dumb terminal, walk to the place you installed it (normally a dark, dusty and warm room, with a lot of green and red Christmas lights), to connect the terminal and to change the settings. What you want is remote management, usually by SNMP, telnet, rlogin or (best) ssh. For all this services you will need a IP. That's the exception to the transparency. The new code allows you without any problem to assign a IP address to the virtual interface formed by the bridge-instance you will create in . All NIC's (or other interfaces) in your bridge will happily listen and respond to datagrams destined to this IP. All other data will not interfere with the bridge. The bridge just acts like a switch. This section describes what you need and how you do to prepare your bridge. Here you can find a list of the files and down-loads you will need for the setup of the bridge. If you have one of the mentioned files or packages on your distribution, of course there is no need to create network load. I'll only mention the files for the 2.2.14 kernel. If you want to try a different one (e.g. 2.2.15 or the recent development kernel) just replace the kernel version number and look whether you find it. You have read the abstract, didn't you? So you know that there is no need to download any kernel-patch if you're working with a kernel later than 2.3.47. Unpatched kernel-sources E.g. linux-2.2.14.tar.bz2 available from your local kernel.org mirror. Please check first if you find it in your distribution (take unpatched kernel-sources). If you don't, please check The Linux Kernel Archive Mirror System for a close by mirror and down-load it from there. Bridge patches If your kernel is later than 2.3.47 you don't need this. The bridging is part of the mainstream from that version. Get the bridge kernel patches for your kernel version from &hbr;. Identify the file by the kernel number. There are also patches allowing to work with IP chains. I never tried it, for I don't see the need to fire-wall inside my LAN, and absolutely no need to bridge against the outer world. Feel free to contribute about that issue. bridge-0.0.9-against-2.2.18.diff, the main kernel patch against 2.2.18 &hbr;patches/bridge-0.0.9-against-2.2.18.diff bridge-ipchains-against-0.0.9-against-2.2.18.diff, an add-on patch for bridge firewalling against 2.2.18 &hbr;patches/bridge-ipchains-against-0.0.9-against-2.2.18.diff bridge-0.0.8-against-2.2.18pre19.diff, the main kernel patch against 2.2.18pre19. &hbr;patches/bridge-0.0.8-against-2.2.18pre19.diff bridge-0.0.8-against-2.2.17-0.5.diff, the main kernel patch against 2.2.17-0.5 &hbr;patches/bridge-0.0.8-against-2.2.17-0.5.diff bridge-ipchains-against-0.0.8-against-2.2.18pre19.diff, an add-on patch for bridge firewalling against 2.2.18pre19 &hbr;patches/bridge-ipchains-against-0.0.8-against-2.2.18pre19.diff bridge-ipchains-against-0.0.8-against-2.2.17-0.5.diff, an add-on patch for bridge firewalling against 2.2.17-0.5 &hbr;patches/bridge-ipchains-against-0.0.8-against-2.2.17-0.5.diff bridge-0.0.7-against-2.2.18pre15.diff, the main kernel patch against 2.2.18pre15 &hbr;patches/bridge-0.0.7-against-2.2.18pre15.diff bridge-ipchains-against-0.0.7-against-2.2.18pre15.diff, an add-on patch for bridge firewalling against 2.2.18pre15 &hbr;patches/bridge-ipchains-against-0.0.7-against-2.2.18pre15.diff bridge-0.0.7-against-2.2.17.diff, the main kernel patch against 2.2.17 &hbr;patches/bridge-0.0.7-against-2.2.17.diff bridge-ipchains-against-0.0.7-against-2.2.17.diff, an add-on patch for bridge firewalling against 2.2.17 &hbr;patches/bridge-ipchains-against-0.0.7-against-2.2.17.diff Bridge configuration utilities You also will need the bridge configuration utilities to set up the bridge . You can also download them from &hbr;. If your kernel is later than 2.3.47 you don't need this. The bridging is part of the mainstream from that version. Apply the bridging patch your kernel. If you don`t know how to do that read the Kernel-HOWTO which can be found in your distribution or at &ldp;HOWTO/Kernel-HOWTO.html root@mbb-1:~ # cd /usr/src/linux-2.2.14 root@mbb-1:/usr/src/linux-2.2.14 # patch -p1 < \ bridge-0.0.5-against-2.2.14.diff . . Now it's time we configure our freshly patched kernel to create the ability to bridge. Run make config, make menuconfig or the click-o-rama make xconfig. Select bridging in the networking option section to be compiled as a module. AFAIK there is no strong reason why not to compile it as a kernel module, whereas I heard rumors about problems with compiling the bridging code directly into the kernel. root@mbb-1:~ # cd /usr/src/linux-2.2.14 root@mbb-1:/usr/src/linux-2.2.14 # make menuconfig . Compile your kernel . Make the new compiled kernel-image to be loaded. I don't know if the kernel patches only apply to the bridging-module or also modify some interfaces inside vmlinuz. So it might not be a error to give a reboot after you updated the kernel-image. root@mbb-1:/usr/src/linux-2.2.14 # make dep clean zImage modules modules_install zlilo ... This is how to compile and install from the scratch. Just unzip the utilities-tarball, cd into the newly created directory and give a make. root@mbb-1:/usr/src/linux-2.2.14 # cd /usr/local/src root@mbb-1:/usr/local/src/ # tar xzvf bridge-utils-0.9.1.tar.gz ..... .... root@mbb-1:/usr/local/src # cd bridge root@mbb-1:/usr/local/src/bridge # make ..... .... After the compilation shown in have worked properly, you can copy the executables to let's say /usr/local/sbin/ (at least I did). So the commands you have to give should be clear, but to be complete see root@mbb-1:/usr/local/src/bridge # cd brctl root@mbb-1:/usr/local/src/bridge/brctl # cp brctl /usr/local/sbin root@mbb-1:/usr/local/src/bridge/brctl # chmod 700 /usr/local/sbin/brctl root@mbb-1:/usr/local/src/bridge/brctl # cp brctld /usr/local/sbin root@mbb-1:/usr/local/src/bridge/brctl # chmod 700 /usr/local/sbin/brctld Also now you can copy the new man-page to a decent place, as shown in . root@mbb-1:/usr/local/src/bridge # cd doc root@mbb-1:/usr/local/src/bridge/doc # gzip -c brctl.8 > /usr/local/man/man8/brctl.8.gz setup Make sure all your network cards are working nicely and are accessible. If so, ifconfig will show you the hardware layout of the network-interface. If you have problems making your cards work please read the Ethernet-HOWTO at &ldp;HOWTO/Ethernet-HOWTO.html. Don't mess around with IP-addresses or net-masks. You will not need it, until you bridge fully operational an up. After you did the steps mentioned above a modprobe -v bridge should show no errors. You can check the success by issuing a cat /proc/modules. Also for each of the network cards you want to use in the bridge the ifconfig whateverNameYourInterfaceHas should give you some information about the interface. If your bridge-utilities have been correctly built and your kernel and bridge-module are OK, then issuing a brctl should show a small command synopsis. root@mbb-1:~ # brctl commands: addbr <bridge> add bridge addif <bridge> <device> add interface to bridge delbr <bridge> delete bridge delif <bridge> <device> delete interface from bridge show show a list of bridges showbr <bridge> show bridge info showmacs <bridge> show a list of mac addrs setageing <bridge> <time> set ageing time setbridgeprio <bridge> <prio> set bridge priority setfd <bridge> <time> set bridge forward delay setgcint <bridge> <time> set garbage collection interval sethello <bridge> <time> set hello time setmaxage <bridge> <time> set max message age setpathcost <bridge> <port> <cost> set path cost setportprio <bridge> <port> <prio> set port priority stp <bridge> <state> {dis,en}able stp The brctl addbr bridgename command creates a logical bridge instance with the name bridgename. You will need at least one logical instance to do any bridging at all. You can interpret the logical bridge being a container for the interfaces taking part in the bridging. Each bridging instance is represented by a new network interface. The corresponding shutdown command is brctl delbr bridgename. brctl delbr bridgename will only work, if there are no more interfaces added to the instance you want to delete. The brctl addif bridgename device command enslaves the network device device to take part in the bridging of bridgename. As a simple explanation, each interface enslaved into a instance is bridged to the other interfaces of the instance. The corresponding command to tale a interface out of the bridge would be brctl delif bridgename device The brctl show command gives you a summary about the overall bridge status, and the instances running as shown in . If you are interested in detailed information about a instance and it's interfaces you will have to check . The brctl showbr bridgename command gives you a summary about a bridge instance and it's enslaved interfaces. The brctl showmacs bridgename command gives you a list of &mac;-addresses of the interfaces which are enslaved in bridgename. Sets the aging time. The aging time is the number of seconds a &mac;-address will be kept in the forwarding database after having received a packet from this &mac; address. The entries in the forwarding database are periodically timed out to ensure they won't stay around forever. Normally there should be no need to modify this parameter. Sets the bridge's relative priority. The bridge with the lowest priority will be elected as the root bridge. The root bridge is the central bridge in the spanning tree. More information about &stp; you find at . Sets the forwarding delay time. The forwarding delay time is the time spent in each of the Listening and Learning states before the Forwarding state is entered. Sets the garbage collection interval. Every (this number) seconds, the entire forwarding database is checked for timed-out entries. The timed-out entries are removed. Sets the hello time. Every (this number) seconds, a hello packet is sent out by the Root Bridge and the Designated Bridges. Hello packets are used to communicate information about the topology throughout the entire Bridged Local Area Network. More information about &stp; you find at . Sets the maximum message age. If the last seen (received) hello packet is more than this number of seconds old, the bridge in question will start the takeover procedure in attempt to become the Root Bridge itself. More information about &stp; you find at . Sets the cost of receiving (or sending, I'm not sure) a packet on this interface. Faster interfaces should have lower path costs. These values are used in the computation of the minimal spanning tree. More information about &stp; you find at . Paths with lower costs are likelier to be used in the spanning tree than high-cost paths (As an example, think of a gigabit line with a 100Mbit or 10Mbit line as a backup line. You don't want the 10/100Mbit line to become the primary line there.) The Linux implementation currently sets the path cost of all eth* interfaces to 100, the nominal cost for a 10Mbit connection. There is unfortunately no easy way to discern 10Mbit from 100Mbit from 1Gbit Ethernet cards, so the bridge cannot use the real interface speed. With this parameter you can enable or disable the Spanning Tree Protocol. This parameters are only of interest, if you have more than one bridge in your LAN and stp enabled. Before modifying them you should read . The standard configuration should consist of: Create the bridge interface. Add interfaces to the bridge. Zero IP the interfaces. Put up the bridge. Optionally you can configure the virtual interface mybridge to take part in your network. It behaves like one interface (like a normal network card). Exactly that way you configure it, replacing the previous command with something like: A more sophisticated setup script you will find at . If you get the terrible experience of a frozen system or some nasty behavior of your nicely shaped linux box at root@mbb-1:~ # ifconfig ethn 0 0.0.0.0 please try (after the reboot of the system if necessary) before starting any bridge stuff at all a root@mbb-1:~ # ifconfig ethn promisc up If again your system is frozen it's you NIC's driver you have to blame, not the bridging code. Here we will cover some advanced features of the new bridge code. STP You are a networkadmin...? You have a switch on top of your ethernet tree...? You have nightmares of a switch emmiting smoke...? Your company is not extremely rich and con provide another redundant switch just waiting for you to plug the patchwires..? You don't feel like placing your bed close to your main network node to plug the wires...? Join linux bridge community and enjoy the relaxment a stp-enabled inhouse scenario is offering to you. Ok, let's leave that commercial and get back linux and the bridge. Take a look on this small thread from the linux-bridge mailing list. Could you give me some hints about multi-bridge scenarios. You can just set up two mirrored bridges. You have two network interfaces in your bridge? Set up the mirror bridge so that it has two network interfaces as well, and connect each of the interfaces to one subnet. This will work without the need of configuration. Be sure that you have the spanning tree protocol enabled. If you didn't use brctl, this should be fine, because in Linux, it is on by default. To check, you could check whether the bridge sends a packet to 0180c2000000 every 2 seconds. If it does, the &stp; is on. The &stp; is needed so that only one bridge will be active at any given time. To be able to see nicely formatted stp packages in your network take a look at the bridge homepage for the patches to tcpdump. The master bridge will send out &stp; packets every 2 seconds by default. The slave bridge will receive these packets, and will notice that the master is still up. If the slave hasn't received a packet in 20 seconds (max. message age parameter), it will start the takeover procedure. From the moment the takeover procedure starts, it will take about 30 seconds (twice the forward delay parameter) for the bridge to become fully operational. Does the &stp; heartbeat mechanism also work with bridges with more than two cards? Yes, it works with any number of interfaces. You can invent bizarre topologies to your heart's desire. You can use multiple (redundant) bridge-bridge connects, you can insert loops, whatever. The &stp; code will always find the minimal spanning tree. The bridge code will even deal with the loss of any number of interfaces. If there are two redundant bridges with identical connections, the loss of an interface on one of the bridges will cause the other bridge to take over forwarding to that specific interface. Now isn't that great? :) How fast does it get up, and what can I do about it? If you think 50 seconds is too much -- and I guess you should; alas, the IEEE specs gives us these default values -- you can tweak these parameters. If you set the hello time (the &stp; packet interval) from 2 to 1 second, you can safely set the message age parameter to 4 seconds. Then you can set the forward delay to 4 seconds, and this will in total give you a takeover time of ~12 seconds. The great thing which is made possible by &stp; is a redundant parallel bridging scenario, with automated take over features. Within a network basing on stp the bridges always try to send a datagram the (by path cost) shortest path. Only on that path the bridges are forwarding, all other paths between this shortest way are blocked. If there is a broken path, the bridges agree about the next shortest. So doubled paths don't break the net, but are bringing more security... For a example setup of a fail secured connection see . The normal idea about a bridge would not allow anything like firewalling, but since several people have asked Lennert for ipchains firewalling on bridge forwarding he implemented it. If you want to do this, you will need to apply the special ip-chain-bridge-patch (also available at the bridge homepage). As soon you have everything up correctly, the bridging code will check each to-be-forwarded packet against the ipchains chain which has the same name as the bridge. So.. if a packet on eth0 is to be forwarded to eth1, and those interfaces are both part of the bridge group br0, the bridging code will check the packet against the chain called 'br0'. If the chain does not exist, the packet will be forwarded. So if you want to do firewalling, you'll have to create the chain yourself. Example: # brctl addbr br0 # brctl addif br0 eth0 # brctl addif br0 eth1 # ifconfig br0 10.0.0.254 # ipchains -N br0 # ipchains -A br0 -s 10.0.0.1/8 -i eth0 -j DENY Creating a bridge interface named br0 Placing eth0 and eth1 into the bridge. Assigning a regular IP address to the bridge. The IP is taken from private network 10.X.X.X (Class A). Creating a ip-chain named br0 It's vital to have the same name here (br0 or whatever you have selected, as long as you have the same in all places). Denying all trafic with source 10.X.X.X on eth0. This is a real-world example which is currently working in our network. Even if it's for sure not a very common situation it might be useful. I had to solve a small hardware incompatibility. HP-VG (Voice Grade) 100Mbit network is not fast Ethernet compatible. Having neither the money nor the will to replace the stuff and having the need to expand the system I had to find a solution which was a) stable and b) cheap. For sure buying a HP modular switch was not meeting condition b). So I remembered I heard about Linux-bridging which automatically fulfilled condition a) and b). So quite some time ago I successfully set up a bridge between the two incompatible networks. Its first hardware-layout is shown in .

Old Bridge Hardware setup The old setup of my previous linux bridge

It was configured as a transparent network component, meaning it didn't take a part in the network, but only bridged it. Originally it was set up on kernel 2.0.35 from a SuSE 5.3 distribution. The next problem showed up at once. A single bridge connecting the big segments might be c) a bottleneck and d) a reason to kill the netadmin, if it blows up. So I tried to find some solution for that problem. What happened next was that I discovered some hints that a new maintainer took over the bridging code. A few mails on the bridge-mailing list later as shown in I was more clever. The new modular bridging code fulfilled exactly what I was looking for. His project page can be found at &hbr; IMHO he's doing a great job. Thanks a lot. The ideas and hints I got from the mailing list discussion shown in lead to a new hardware-setup shown in . The setup is intended to provide a default machine (guess which one). The bridge has 3 HP cards of which each is connected to a HP VG15 hub. The 3com card is connected to a 3com Superstack Fast Ethernet switch.

Multi bridge Hardware Setup The practically working setup of my local linux Ethernet multi bridge

This setup is not only fail proof to any one of the bridge's interfaces being down, but also to complete blackout of one of the bridges. Additional advantage to the old-setup that the single HUBS are switched. This means that a datagram being sent from one port on the VG15 HUB blocks 30 ports by maximum and 15 ports by minimum, instead of blocking all 45 ports. Also, the breakdown of the HUB, to the old bridge was connected, would have caused the whole HP-segment to break down. With the new code only the machines connected to the broken HUB will get no more data. For both bridges the setup is exactly the same (with the exception of bridge priority which will be discussed later on). The machine was setup by the SuSE 6.4 distribution with the original unpatched kernel sources installed. At this point only the minimal configuration and no additional hardware or network setup. The basic setup is according the descriptions in the beginning of this document. The thing I did in addition was bringing up the unpatched 2.2.14 sources of the SuSE 6.4 distribution to version 2.2.15 as in . root@mbb-1:~ # cd /usr/src/linux-2.2.14 root@mbb-1:/usr/src/linux-2.2.14 # patch -p1 \ /usr/local/download/kernel/patch-2.2.15 patching file ........................ patching file ................... ... .. root@mbb-1:/usr/src/linux-2.2.14 # cd .. root@mbb-1:/usr/src # mv linux-2.2.14 linux-2.2.15 root@mbb-1:/usr/src # rm linux root@mbb-1:/usr/src # ln -s linux-2.2.15 linux Next step was to apply the bridge-patch as shown in . root@mbb-1:/usr/src # cd /usr/src/linux-2.2.15 root@mbb-1:/usr/src/linux-2.2.15 # patch -p1 < \ bridge-0.0.5-against-2.2.15.diff patching file ........................ patching file ................... ... .. After that I selected the bridging code to be compiled as a module as shown in . root@mbb-1:/usr/src/linux-2.2.15 # make config .. * * Code maturity level options * Prompt for development and/or incomplete code/drivers (CONFIG_EXPERIMENTAL) [N/y/?] Y .. 802.1d Ethernet Bridging (CONFIG_BRIDGE) [N/y/m/?] (NEW) m .. By the way I also selected the drivers of my NIC's to be compiled as modules which resulted to 3c95x.o and hp100.o. root@mbb-1:/usr/src/linux-2.2.15 # make dep clean zImage \ modules modules_install zlilo .. root@mbb-1:/usr/src/linux-2.2.15 # init 6 After the reboot happening I started at runlevel 1 leaving all the networking out of the running system. That gave me the chance to check the setup step by step. The command modprobe -v bridge worked without any warnings, so that one was OK. Next I edited my /etc/modules.conf by aliasing my network card drivers as shown in and . I didn't need to make use of the options, all cards where realized proper as I checked by cat /proc/modules, cat /proc/interrupts and cat /proc/ioports. # Aliases - specify your hardware alias eth0 3c59x alias eth1 hp100 alias eth2 hp100 alias eth3 hp100 # Aliases - specify your hardware alias eth0 3c509 alias eth1 hp100 alias eth2 hp100 alias eth3 hp100 So next thing would have been a step by step setup of the bridge and it's interfaces. Because I'm lazy I just show the init script I prepared for the setup. Of course you'll have do adapt the script to your system, if you want to use it. Please remember I'm writing this for the setup of a SuSE distribution. #! /bin/bash # Copyright (c) 2000 Uwe Böhme. All rights reserved. # # Author: Uwe Böhme <uwe@bnhof.de>, 2000 # # # /sbin/init.d/bridge # . /etc/rc.config return=$rc_done case "$1" in start) echo "Starting service bridge mueb" brctl addbr mueb || return=$rc_failed brctl setbridgeprio mueb 0 || return=$rc_failed brctl addif mueb eth0 || return=$rc_failed brctl addif mueb eth1 || return=$rc_failed brctl addif mueb eth2 || return=$rc_failed brctl addif mueb eth3 || return=$rc_failed ifconfig eth0 0.0.0.0 || return=$rc_failed ifconfig eth1 0.0.0.0 || return=$rc_failed ifconfig eth2 0.0.0.0 || return=$rc_failed ifconfig eth3 0.0.0.0 || return=$rc_failed brctl sethello mueb 1 || return=$rc_failed brctl setmaxage mueb 4 || return=$rc_failed brctl setfd mueb 4 || return=$rc_failed echo -e "$return" ;; stop) echo "Shutting down service bridge mueb" brctl delif mueb eth3 || return=$rc_failed brctl delif mueb eth2 || return=$rc_failed brctl delif mueb eth1 || return=$rc_failed brctl delif mueb eth0 || return=$rc_failed brctl delbr mueb || return=$rc_failed rmmod bridge || return=$rc_failed echo -e "$return" ;; status) ifconfig mueb brctl showbr mueb ;; restart) $0 stop && $0 start || return=$rc_failed ;; *) echo "Usage: $0 {start|stop|status|restart}" exit 1 esac test "$return" = "$rc_done" || exit 1 exit 0 This command creates a new virtual interface (bridge instance) with the name mueb and also brings up the bridge module. At least my system it does. Maybe you have to enable the kernel module loader. Here the script sets the bridge's priority (relative to other bridges in the net) to 0. This is indicating that this bridge will become the root bridge as long as there is no other bridge with a lower priority level available. In the init script of the backup bridge this line in missing, leaving it with the default priority of 100. Enslaves the Ethernet interface to become a port in the bridge. Takes away any possibly disturbing IP-address and brings the interface up. Setting the hello time of the bridge to one second makes it possible to reduce the maxage value of the bridges inside the network. Setting the time the a bridge is waiting before starting the takeover process to a shorter period. Forcing the bridge to forward earlier than the default time. Take the Ethernet out of the bridging instance. Destroy the bridge instance. Remove the bridge module. To polish your setup and to be able to reach the bridge from remote you now can configure your bridge instance as if it would be a physical existing network interface. You can give it a nice IP with a suitable net-mask. It doesn't matter from which segment in you net, you will reach the bridge with this IP-address. Here I want to show and explain about how the running bridge shows up. The output of bridge@mbb-1 is the output of the primary bridge, while you see in the output of the backup bridge waiting to take over. mueb bridge id 0000.0800062815f6 designated root 0000.0800062815f6 root port 0 path cost 0 max age 4.00 bridge max age 4.00 hello time 1.00 bridge hello time 1.00 forward delay 4.00 bridge forward delay 4.00 ageing time 300.00 gc interval 4.00 hello timer 0.80 tcn timer 0.00 topology change timer 0.00 gc timer 3.80 flags eth0 (1) port id 8001 state forwarding designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.00 designated port 8001 forward delay timer 0.00 designated cost 0 hold timer 0.80 flags eth1 (2) port id 8002 state forwarding designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.00 designated port 8002 forward delay timer 0.00 designated cost 0 hold timer 0.80 flags eth2 (3) port id 8003 state forwarding designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.00 designated port 8003 forward delay timer 0.00 designated cost 0 hold timer 0.80 flags eth3 (4) port id 8004 state forwarding designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.00 designated port 8004 forward delay timer 0.00 designated cost 0 hold timer 0.80 flags mueb bridge id 0064.00a024d04cd6 designated root 0000.0800062815f6 root port 1 path cost 100 max age 4.00 bridge max age 4.00 hello time 1.00 bridge hello time 1.00 forward delay 4.00 bridge forward delay 4.00 ageing time 300.00 gc interval 4.00 hello timer 0.00 tcn timer 0.00 topology change timer 0.00 gc timer 2.39 flags eth0 (1) port id 8001 state forwarding designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.42 designated port 8001 forward delay timer 0.00 designated cost 0 hold timer 0.00 flags eth1 (2) port id 8002 state blocking designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.42 designated port 8002 forward delay timer 0.00 designated cost 0 hold timer 0.00 flags eth2 (3) port id 8003 state blocking designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.42 designated port 8003 forward delay timer 0.00 designated cost 0 hold timer 0.00 flags eth3 (4) port id 8004 state blocking designated root 0000.0800062815f6 path cost 100 designated bridge 0000.0800062815f6 message age timer 0.42 designated port 8004 forward delay timer 0.00 designated cost 0 hold timer 0.00 flags If you take a glance into /var/log/messages as shown in and in you can see how the bridges are coming up and deciding how to do their duty. mbb-1 has a lower value for bridge-priority (see ), telling it to try to become the root bridge. As you can see mbb-1 forwards all ports, while mbb-2 blocks all ports with the exception of eth0. May 25 16:46:04 mbb-1 init: Switching to runlevel: 2 May 25 16:46:04 mbb-1 kernel: NET4: Ethernet Bridge 008 for NET4.0 May 25 16:46:04 mbb-1 kernel: device eth0 entered promiscuous mode May 25 16:46:04 mbb-1 kernel: device eth1 entered promiscuous mode May 25 16:46:04 mbb-1 kernel: device eth2 entered promiscuous mode May 25 16:46:04 mbb-1 kernel: device eth3 entered promiscuous mode May 25 16:46:04 mbb-1 kernel: mueb: port 4(eth3) entering listening state May 25 16:46:04 mbb-1 kernel: mueb: port 3(eth2) entering listening state May 25 16:46:04 mbb-1 kernel: mueb: port 2(eth1) entering listening state May 25 16:46:04 mbb-1 kernel: mueb: port 1(eth0) entering listening state May 25 16:46:08 mbb-1 kernel: mueb: port 4(eth3) entering learning state May 25 16:46:08 mbb-1 kernel: mueb: port 3(eth2) entering learning state May 25 16:46:08 mbb-1 kernel: mueb: port 2(eth1) entering learning state May 25 16:46:08 mbb-1 kernel: mueb: port 1(eth0) entering learning state May 25 16:46:12 mbb-1 kernel: mueb: port 4(eth3) entering forwarding state May 25 16:46:12 mbb-1 kernel: mueb: topology change detected, propagating May 25 16:46:12 mbb-1 kernel: mueb: port 3(eth2) entering forwarding state May 25 16:46:12 mbb-1 kernel: mueb: topology change detected, propagating May 25 16:46:12 mbb-1 kernel: mueb: port 2(eth1) entering forwarding state May 25 16:46:12 mbb-1 kernel: mueb: topology change detected, propagating May 25 16:46:12 mbb-1 kernel: mueb: port 1(eth0) entering forwarding state May 25 16:46:12 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:06:16 mbb-2 init: Switching to runlevel: 2 Jun 8 06:06:17 mbb-2 kernel: NET4: Ethernet Bridge 008 for NET4.0 Jun 8 06:06:17 mbb-2 kernel: device eth0 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth1 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth2 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth3 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: mueb: port 4(eth3) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 3(eth2) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 1(eth0) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 2(eth1) entering blocking state Jun 8 06:06:17 mbb-2 kernel: mueb: port 3(eth2) entering blocking state Jun 8 06:06:17 mbb-2 kernel: mueb: port 4(eth3) entering blocking state Jun 8 06:06:21 mbb-2 kernel: mueb: port 1(eth0) entering learning state Jun 8 06:06:25 mbb-2 kernel: mueb: port 1(eth0) entering forwarding state To check if really all the promised features are working, I did some crude test. The message logs are shown here in. I think just taking a patch wire out of a bridge port is a really good real survival test. So I pulled the plugs one by one out of the sockets and looked what happened. To give you not too much tension let me summarize first: It's really working. All the takeovers happened within less then 12 seconds. The really interesting messages you can find at mbb-2. To see how everything comes up, I stopped network services first. In you will see the messages caused by a init 2 followed by a take out the plug, wait what happens, then place it back in the order eth3, eth2, eth1, eth0 . The thing I did, was making the tests, and publishing the dump. The one writing the nice explanations was Lennert again. Jun 8 06:06:16 mbb-2 init: Switching to runlevel: 2 Jun 8 06:06:17 mbb-2 kernel: NET4: Ethernet Bridge 008 for NET4.0 Jun 8 06:06:17 mbb-2 kernel: device eth0 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth1 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth2 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: device eth3 entered promiscuous mode Jun 8 06:06:17 mbb-2 kernel: mueb: port 4(eth3) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 3(eth2) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 1(eth0) entering listening state Jun 8 06:06:17 mbb-2 kernel: mueb: port 2(eth1) entering blocking state Jun 8 06:06:17 mbb-2 kernel: mueb: port 3(eth2) entering blocking state Jun 8 06:06:17 mbb-2 kernel: mueb: port 4(eth3) entering blocking state Jun 8 06:06:21 mbb-2 kernel: mueb: port 1(eth0) entering learning state Jun 8 06:06:25 mbb-2 kernel: mueb: port 1(eth0) entering forwarding state Jun 8 06:07:15 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 4(eth3) Jun 8 06:07:15 mbb-2 kernel: mueb: port 4(eth3) entering listening state Jun 8 06:07:19 mbb-2 kernel: mueb: port 4(eth3) entering learning state Jun 8 06:07:23 mbb-2 kernel: mueb: port 4(eth3) entering forwarding state Jun 8 06:07:23 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:08:51 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:08:51 mbb-2 kernel: mueb: port 4(eth3) entering blocking state Jun 8 06:09:22 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 3(eth2) Jun 8 06:09:22 mbb-2 kernel: mueb: port 3(eth2) entering listening state Jun 8 06:09:26 mbb-2 kernel: mueb: port 3(eth2) entering learning state Jun 8 06:09:30 mbb-2 kernel: mueb: port 3(eth2) entering forwarding state Jun 8 06:09:30 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:10:09 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:10:09 mbb-2 kernel: mueb: port 3(eth2) entering blocking state Jun 8 06:10:10 mbb-2 kernel: mueb: retransmitting tcn bpdu Jun 8 06:10:41 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 2(eth1) Jun 8 06:10:41 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 8 06:10:45 mbb-2 kernel: mueb: port 2(eth1) entering learning state Jun 8 06:10:49 mbb-2 kernel: mueb: port 2(eth1) entering forwarding state Jun 8 06:10:49 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:11:06 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:11:06 mbb-2 kernel: mueb: port 2(eth1) entering blocking state Jun 8 06:11:33 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 1(eth0) Jun 8 06:11:33 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 8 06:11:37 mbb-2 kernel: mueb: port 2(eth1) entering learning state Jun 8 06:11:41 mbb-2 kernel: mueb: port 2(eth1) entering forwarding state Jun 8 06:11:41 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:14:18 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 8 06:14:18 mbb-2 kernel: mueb: port 2(eth1) entering blocking state Jun 8 06:14:19 mbb-2 kernel: mueb: retransmitting tcn bpdu The kernel sees that there are already bridges (actually, only one of them, but Hello packets are coming in on all 4 of the ports) on eth[0123]. To maintain connectivity with the rest of the network, the bridge decides to keep port 1 (eth0) active (i.e. in the forwarding state), and to temporarily disable ports 2-4. The plug on eth3 was pulled. Here you can see that the message age timer expired (). The last Hello packet was seen more than X seconds ago. The bridge concludes that the connection to the bridge that was there has died. Therefore, it is going to try to enable this port, to provide network connectivity to the now-cutoff segment. It enters the listening state. It waits to see whether the old bridge might come back, or whether another bridge is going to claim takeover. Okay, no other bridge was seen. We're going to try to provide network connectivity to this segment ourselves. Which means: we're going to try and become designated bridge for this segment. We now enter the learning state. In this state, we only learn &mac; addresses and we do not forward yet. This is because if we see an unknown destination address, we send the datagram to all ports, and this flooding will happen unnecessarily often if we have a empty &mac; table. Therefore, we're going to fill up our &mac; table with useful entries first, and this is what happens during the learning state. Okay, here we go. Pray for us. Because we took over for this segment, all communication towards this segment now goes through this bridge. This means that the topology has changed. If the topology changes, we must let all bridges now, so that they can time out stale &mac; address location data quickly. This is why we send Topology Change Notification Bridge Protocol Data Units (tcn bpdus). Apparently the root bridge immediately acknowledges this tcn bpdu in the next Hello message it sends (the protocol requires for the root bridge to acknowledge it), because this is the only such message we see. In situations where you see loads of these messages, it means that the root bridge cannot acknowledge them, which probably means your root bridge has a twisted STP implementation. Hey, something happened again! Yup... eth3 came back online. The root bridge will provide connectivity for this segment again, so that we can disable this port. Same story for eth2, eth1 and eth0. This means the tcn bpdu wasn't acknowledged quick enough. That is why it is retransmitted. The root bridge mbb-1 was not so chatty. It only reported some topology changes and propagated them as you can see in . If somebody can offer a explanation why the root bridge is so quiet in messaging please tell me. Jun 8 06:06:52 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:06:52 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:07:31 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:07:31 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:07:32 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:07:32 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:08:11 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:08:11 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:08:29 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:08:29 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:09:03 mbb-1 kernel: mueb: received tcn bpdu on port 2(eth1) Jun 8 06:09:03 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:11:40 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:11:40 mbb-1 kernel: mueb: topology change detected, propagating Jun 8 06:11:41 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 8 06:11:41 mbb-1 kernel: mueb: topology change detected, propagating One of the other bridges tells us that the topology of the LAN has changed (see ). Well, okay. We will set lower timeouts on our &mac;C table for a short period of time, and we will propagate this topology change throughout the network. The ultimate test is of course a total blocking, breakdown or something similar to the root bridge. I did this by shooting down the root bridge by init 1. Next I brought it up again with init 2. Last I pulled all plugs out of the root bridge and waited for some time before I placed them again. In you will see the messages from the master-bridge mbb-1, and in you see what happened the same time at the backup-bridge mbb-2. Jun 12 13:35:15 mbb-1 init: Switching to runlevel: 1 Jun 12 13:35:20 mbb-1 kernel: mueb: port 4(eth3) entering disabled state Jun 12 13:35:20 mbb-1 kernel: mueb: port 3(eth2) entering disabled state Jun 12 13:35:20 mbb-1 kernel: mueb: port 2(eth1) entering disabled state Jun 12 13:35:20 mbb-1 kernel: mueb: port 1(eth0) entering disabled state Jun 12 13:35:20 mbb-1 kernel: mueb: port 2(eth1) entering disabled state Jun 12 13:35:20 mbb-1 kernel: device eth1 left promiscuous mode Jun 12 13:35:20 mbb-1 kernel: mueb: port 1(eth0) entering disabled state Jun 12 13:35:20 mbb-1 kernel: device eth0 left promiscuous mode Jun 12 13:35:20 mbb-1 kernel: mueb: port 4(eth3) entering disabled state Jun 12 13:35:20 mbb-1 kernel: device eth3 left promiscuous mode Jun 12 13:35:20 mbb-1 kernel: mueb: port 3(eth2) entering disabled state Jun 12 13:35:20 mbb-1 kernel: device eth2 left promiscuous mode Jun 12 13:35:50 mbb-1 init: Switching to runlevel: 2 Jun 12 13:35:50 mbb-1 kernel: NET4: Ethernet Bridge 008 for NET4.0 Jun 12 13:35:51 mbb-1 kernel: device eth0 entered promiscuous mode Jun 12 13:35:51 mbb-1 kernel: device eth1 entered promiscuous mode Jun 12 13:35:51 mbb-1 kernel: device eth2 entered promiscuous mode Jun 12 13:35:51 mbb-1 kernel: device eth3 entered promiscuous mode Jun 12 13:35:51 mbb-1 kernel: mueb: port 4(eth3) entering listening state Jun 12 13:35:51 mbb-1 kernel: mueb: port 3(eth2) entering listening state Jun 12 13:35:51 mbb-1 kernel: mueb: port 2(eth1) entering listening state Jun 12 13:35:51 mbb-1 kernel: mueb: port 1(eth0) entering listening state Jun 12 13:35:51 mbb-1 kernel: mueb: received tcn bpdu on port 2(eth1) Jun 12 13:35:51 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:52 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 12 13:35:52 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:55 mbb-1 kernel: mueb: port 4(eth3) entering learning state Jun 12 13:35:55 mbb-1 kernel: mueb: port 3(eth2) entering learning state Jun 12 13:35:55 mbb-1 kernel: mueb: port 2(eth1) entering learning state Jun 12 13:35:55 mbb-1 kernel: mueb: port 1(eth0) entering learning state Jun 12 13:35:59 mbb-1 kernel: mueb: port 4(eth3) entering forwarding state Jun 12 13:35:59 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:59 mbb-1 kernel: mueb: port 3(eth2) entering forwarding state Jun 12 13:35:59 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:59 mbb-1 kernel: mueb: port 2(eth1) entering forwarding state Jun 12 13:35:59 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:59 mbb-1 kernel: mueb: port 1(eth0) entering forwarding state Jun 12 13:35:59 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:39:03 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 12 13:39:03 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:39:05 mbb-1 kernel: mueb: received tcn bpdu on port 1(eth0) Jun 12 13:39:05 mbb-1 kernel: mueb: topology change detected, propagating Jun 12 13:35:21 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 4(eth3) Jun 12 13:35:21 mbb-2 kernel: mueb: port 4(eth3) entering listening state Jun 12 13:35:21 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 3(eth2) Jun 12 13:35:21 mbb-2 kernel: mueb: port 3(eth2) entering listening state Jun 12 13:35:21 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 2(eth1) Jun 12 13:35:21 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 12 13:35:21 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 1(eth0) Jun 12 13:35:21 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:35:25 mbb-2 kernel: mueb: port 4(eth3) entering learning state Jun 12 13:35:25 mbb-2 kernel: mueb: port 3(eth2) entering learning state Jun 12 13:35:25 mbb-2 kernel: mueb: port 2(eth1) entering learning state Jun 12 13:35:29 mbb-2 kernel: mueb: port 4(eth3) entering forwarding state Jun 12 13:35:29 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:35:29 mbb-2 kernel: mueb: port 3(eth2) entering forwarding state Jun 12 13:35:29 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:35:29 mbb-2 kernel: mueb: port 2(eth1) entering forwarding state Jun 12 13:35:29 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:35:49 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 12 13:35:49 mbb-2 kernel: mueb: port 3(eth2) entering blocking state Jun 12 13:35:49 mbb-2 kernel: mueb: topology change detected, \ <6>mueb: port 4(eth3) entering blocking state Jun 12 13:35:49 mbb-2 kernel: mueb: topology change detected, \ <6>mueb: port 2(eth1) entering blocking state Jun 12 13:35:50 mbb-2 kernel: mueb: retransmitting tcn bpdu Jun 12 13:38:26 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 2(eth1) Jun 12 13:38:26 mbb-2 kernel: mueb: port 2(eth1) entering listening state Jun 12 13:38:27 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 3(eth2) Jun 12 13:38:27 mbb-2 kernel: mueb: port 3(eth2) entering listening state Jun 12 13:38:28 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 4(eth3) Jun 12 13:38:28 mbb-2 kernel: mueb: port 4(eth3) entering listening state Jun 12 13:38:30 mbb-2 kernel: mueb: port 2(eth1) entering learning state Jun 12 13:38:30 mbb-2 kernel: mueb: neighbour 0000.08:00:06:28:15:f6 lost on port 1(eth0) Jun 12 13:38:30 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:38:31 mbb-2 kernel: mueb: port 3(eth2) entering learning state Jun 12 13:38:32 mbb-2 kernel: mueb: port 4(eth3) entering learning state Jun 12 13:38:34 mbb-2 kernel: mueb: port 2(eth1) entering forwarding state Jun 12 13:38:34 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:38:35 mbb-2 kernel: mueb: port 3(eth2) entering forwarding state Jun 12 13:38:35 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:38:36 mbb-2 kernel: mueb: port 4(eth3) entering forwarding state Jun 12 13:38:36 mbb-2 kernel: mueb: topology change detected, propagating Jun 12 13:39:01 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 12 13:39:01 mbb-2 kernel: mueb: port 3(eth2) entering blocking state Jun 12 13:39:01 mbb-2 kernel: mueb: topology change detected, \ <6>mueb: port 4(eth3) entering blocking state Jun 12 13:39:02 mbb-2 kernel: mueb: topology change detected, sending tcn bpdu Jun 12 13:39:02 mbb-2 kernel: mueb: port 2(eth1) entering blocking state In this section you will find a (for now) very incomplete list of NIC's which are known to work or known to cause problem. For I neither have the money to buy a lot of different NIC's, nor I have any connections to hardware vendors, I depend on your feedback to keep the list accurate. So feel free to mail about success or failure to Uwe Böhme. - - - Cards I tried and are also reported not to work by other people - - Cards I tried or are reported not to work by other people - Cards reported not to work by other people + Cards reported to work by other people + + Cards I tried or are reported to work by other people + + + Cards I tried and are also reported to work by other people 3c509b Etherlink III + + 3c905b/3c905c + + + Never heard about any problem HP J2585A - - System hang-up after ifconfig, unable to run promisc mode HP J2585B + + AMD PCnet32 10/100 + + RTL (Realtek) 8029 + + Here you will some recommendations which documents you should read before you start to setup a bridge. The bridge home-page Will give you recent information about the bridging code and the bridge utilities. &ldp;HOWTO/NET3-4-HOWTO Describes how to install and configure the Linux networking software and associated tools. &ldp;/HOWTO/Ethernet-HOWTO Information about which Ethernet devices can be used for Linux, and how to set them up (focused on the hardware and low level driver aspect of the Ethernet cards). Here you will find some of the frequently asked questions connected to bridging. What hardware do I need to run a bridge with 2-n NICs. I think a fat 486 or a modest Pentium should be able to keep up with 2x100Mbit pretty well, but I have never tested this. I don't think RAM will matter much (8 or 16MB and all should be fine). CPU will not matter a whole lot either (486/Pentium and all should be fine). I think the primary contributor is the type of bus (ISA, PCI) and the type of network cards (some network cards require less work than others). Big switches usually have immensely fat internal buses (3 or 4 gigabits is not uncommon). Standard PCI, for example, can't keep up with a gigabit ethernet cards. Can you please recommend some tools to measure a 2-port linux bridge throughput. Well, first question is: does it have 100mbit interfaces? If it hasn't (10mbit only), it shouldn't have problems with keeping up, almost regardless of the processor speed. If it does have 100mbit interfaces and you're not sure it will keep up, you can run a flood ping with big packets across it (ping -f -s 1450 ipaddress) and see whether it keeps up. I'm running with kernel x.x.x. Is a patch out there, to give me chance to use this stuff? There are patches for and 2.2.14, 2.2.15. Since 2.3.47 it's in the mainstream kernel, so you don't need to patch. If you're talking about others, you will have to upgrade, if you need to bridge. I've heared unconfirmed roumors about the 2.2.15 patches working without any change also with the 2.2.16 kernel. Anyone mind telling me about it?