1 <?xml version='1.0' encoding='UTF-8'?>
2 <!-- This document was created with Syntext Serna Free. --><chapter xmlns="http://docbook.org/ns/docbook" xmlns:xl="http://www.w3.org/1999/xlink" version="5.0" xml:lang="en-US" xml:id="settingupbonding">
3 <title xml:id="settingupbonding.title">Setting Up Network Interface Bonding</title>
4 <para>This chapter describes how to use multiple network interfaces in parallel to increase bandwidth and/or redundancy. Topics include:</para>
8 <xref linkend="dbdoclet.50438258_95824"/>
13 <xref linkend="dbdoclet.50438258_29107"/>
18 <xref linkend="dbdoclet.50438258_25353"/>
23 <xref linkend="dbdoclet.50438258_26899"/>
28 <xref linkend="dbdoclet.50438258_54769"/>
33 <xref linkend="dbdoclet.50438258_92244"/>
38 <para>Using network interface bonding is optional.</para>
40 <section xml:id="dbdoclet.50438258_95824">
41 <title>Network Interface Bonding Overview</title>
42 <para>Bonding, also known as link aggregation, trunking and port trunking, is a method of aggregating multiple physical network links into a single logical link for increased bandwidth.</para>
43 <para>Several different types of bonding are available in Linux. All these types are referred to as 'modes', and use the bonding kernel module.</para>
44 <para>Modes 0 to 3 allow load balancing and fault tolerance by using multiple interfaces. Mode 4 aggregates a group of interfaces into a single virtual interface where all members of the group share the same speed and duplex settings. This mode is described under IEEE spec 802.3ad, and it is referred to as either 'mode 4' or '802.3ad.'</para>
46 <section xml:id="dbdoclet.50438258_29107">
47 <title>Requirements</title>
48 <para>The most basic requirement for successful bonding is that both endpoints of the connection must be capable of bonding. In a normal case, the non-server endpoint is a switch. (Two systems connected via crossover cables can also use bonding.) Any switch used must explicitly handle 802.3ad Dynamic Link Aggregation.</para>
49 <para>The kernel must also be configured with bonding. All supported Lustre kernels have bonding functionality. The network driver for the interfaces to be bonded must have the ethtool functionality to determine slave speed and duplex settings. All recent network drivers implement it.</para>
50 <para>To verify that your interface works with ethtool, run:</para>
51 <screen># which ethtool
56 Supported ports: [ TP MII ]
57 Supported link modes: 10baseT/Half 10baseT/Full
58 100baseT/Half 100baseT/Full
59 Supports auto-negotiation: Yes
60 Advertised link modes: 10baseT/Half 10baseT/Full
61 100baseT/Half 100baseT/Full
62 Advertised auto-negotiation: Yes
69 Supports Wake-on: pumbg
71 Current message level: 0x00000001 (1)
77 Supported ports: [ TP MII ]
78 Supported link modes: 10baseT/Half 10baseT/Full
79 100baseT/Half 100baseT/Full
80 Supports auto-negotiation: Yes
81 Advertised link modes: 10baseT/Half 10baseT/Full
82 100baseT/Half 100baseT/Full
83 Advertised auto-negotiation: Yes
90 Supports Wake-on: pumbg
92 Current message level: 0x00000007 (7)
94 To quickly check whether your kernel supports bonding, run:
95 # grep ifenslave /sbin/ifup
100 <section xml:id="dbdoclet.50438258_25353">
101 <title>Bonding Module Parameters</title>
102 <para>Bonding module parameters control various aspects of bonding.</para>
103 <para>Outgoing traffic is mapped across the slave interfaces according to the transmit hash policy. For Lustre, we recommend that you set the <literal>xmit_hash_policy</literal> option to the layer3+4 option for bonding. This policy uses upper layer protocol information if available to generate the hash. This allows traffic to a particular network peer to span multiple slaves, although a single connection does not span multiple slaves.</para>
104 <screen>$ xmit_hash_policy=layer3+4</screen>
105 <para>The miimon option enables users to monitor the link status. (The parameter is a time interval in milliseconds.) It makes an interface failure transparent to avoid serious network degradation during link failures. A reasonable default setting is 100 milliseconds; run:</para>
106 <screen>$ miimon=100</screen>
107 <para>For a busy network, increase the timeout.</para>
109 <section xml:id="dbdoclet.50438258_26899">
110 <title>Setting Up Bonding</title>
111 <para>To set up bonding:</para>
114 <para>Create a virtual 'bond' interface by creating a configuration file in:</para>
115 <screen>/etc/sysconfig/network-scripts/ # vi /etc/sysconfig/ network-scripts/ifcfg-\
119 <para>Append the following lines to the file.</para>
121 IPADDR=192.168.10.79 # Use the free IP Address of your network
123 NETMASK=255.255.255.0
129 <para>Attach one or more slave interfaces to the bond interface. Modify the eth0 and eth1 configuration files (using a VI text editor).</para>
132 <para>Use the VI text editor to open the eth0 configuration file.</para>
133 <screen># vi /etc/sysconfig/network-scripts/ifcfg-eth0</screen>
136 <para>Modify/append the eth0 file as follows:</para>
142 BOOTPROTO=none</screen>
145 <para>Use the VI text editor to open the eth1 configuration file.</para>
146 <screen># vi /etc/sysconfig/network-scripts/ifcfg-eth1</screen>
149 <para>Modify/append the eth1 file as follows:</para>
161 <para>Set up the bond interface and its options in <literal>/etc/modprobe.d/bond.conf</literal>. Start the slave interfaces by your normal network method.</para>
162 <screen># vi /etc/modprobe.d/bond.conf
166 <para>Append the following lines to the file.</para>
167 <screen>alias bond0 bonding
168 options bond0 mode=balance-alb miimon=100
172 <para>Load the bonding module.</para>
173 <screen># modprobe bonding
175 # ifenslave bond0 eth0 eth1
181 <para>Start/restart the slave interfaces (using your normal network method).</para>
183 <para>You must <literal>modprobe</literal> the bonding module for each bonded interface. If you wish to create bond0 and bond1, two entries in <literal>bond.conf</literal> file are required.</para>
185 <para>The examples below are from RedHat systems. For setup use: <literal>/etc/sysconfig/networking-scripts/ifcfg-*</literal> The website referenced below includes detailed instructions for other configuration methods, instructions to use DHCP with bonding, and other setup details. We strongly recommend you use this website.</para>
186 <para><link xl:href="http://www.linuxfoundation.org/collaborate/workgroups/networking/bonding">http://www.linuxfoundation.org/collaborate/workgroups/networking/bonding</link></para>
189 <para>Check /proc/net/bonding to determine status on bonding. There should be a file there for each bond interface.</para>
190 <screen># cat /proc/net/bonding/bond0
191 Ethernet Channel Bonding Driver: v3.0.3 (March 23, 2006)
193 Bonding Mode: load balancing (round-robin)
195 MII Polling Interval (ms): 0
199 Slave Interface: eth0
201 Link Failure Count: 0
202 Permanent HW addr: 4c:00:10:ac:61:e0
204 Slave Interface: eth1
206 Link Failure Count: 0
207 Permanent HW addr: 00:14:2a:7c:40:1d
211 <para>Use ethtool or ifconfig to check the interface state. ifconfig lists the first bonded interface as 'bond0.'</para>
213 bond0 Link encap:Ethernet HWaddr 4C:00:10:AC:61:E0
214 inet addr:192.168.10.79 Bcast:192.168.10.255 \ Mask:255.255.255.0
215 inet6 addr: fe80::4e00:10ff:feac:61e0/64 Scope:Link
216 UP BROADCAST RUNNING MASTER MULTICAST MTU:1500 Metric:1
217 RX packets:3091 errors:0 dropped:0 overruns:0 frame:0
218 TX packets:880 errors:0 dropped:0 overruns:0 carrier:0
219 collisions:0 txqueuelen:0
220 RX bytes:314203 (306.8 KiB) TX bytes:129834 (126.7 KiB)
222 eth0 Link encap:Ethernet HWaddr 4C:00:10:AC:61:E0
223 inet6 addr: fe80::4e00:10ff:feac:61e0/64 Scope:Link
224 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
225 RX packets:1581 errors:0 dropped:0 overruns:0 frame:0
226 TX packets:448 errors:0 dropped:0 overruns:0 carrier:0
227 collisions:0 txqueuelen:1000
228 RX bytes:162084 (158.2 KiB) TX bytes:67245 (65.6 KiB)
229 Interrupt:193 Base address:0x8c00
231 eth1 Link encap:Ethernet HWaddr 4C:00:10:AC:61:E0
232 inet6 addr: fe80::4e00:10ff:feac:61e0/64 Scope:Link
233 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
234 RX packets:1513 errors:0 dropped:0 overruns:0 frame:0
235 TX packets:444 errors:0 dropped:0 overruns:0 carrier:0
236 collisions:0 txqueuelen:1000
237 RX bytes:152299 (148.7 KiB) TX bytes:64517 (63.0 KiB)
238 Interrupt:185 Base address:0x6000
243 <title>Examples</title>
244 <para>This is an example showing <literal>bond.conf</literal> entries for bonding Ethernet interfaces <literal>eth1</literal> and <literal>eth2</literal> to <literal>bond0</literal>:</para>
245 <screen># cat /etc/modprobe.d/bond.conf
249 options bond0 mode=balance-alb miimon=100
251 # cat /etc/sysconfig/network-scripts/ifcfg-bond0
254 NETMASK=255.255.255.0
255 IPADDR=192.168.10.79 # (Assign here the IP of the bonded interface.)
260 # cat /etc/sysconfig/network-scripts/ifcfg-eth0
263 HWADDR=4c:00:10:ac:61:e0
272 <para>In the following example, the <literal>bond0</literal> interface is the master (MASTER) while <literal>eth0</literal> and <literal>eth1</literal> are slaves (SLAVE).</para>
274 <para>All slaves of <literal>bond0</literal> have the same MAC address (Hwaddr) - <literal>bond0</literal>. All modes, except TLB and ALB, have this MAC address. TLB and ALB require a unique MAC address for each slave.</para>
276 <screen>$ /sbin/ifconfig
278 bond0Link encap:EthernetHwaddr 00:C0:F0:1F:37:B4
279 inet addr:XXX.XXX.XXX.YYY Bcast:XXX.XXX.XXX.255 Mask:255.255.252.0
280 UP BROADCAST RUNNING MASTER MULTICAST MTU:1500 Metric:1
281 RX packets:7224794 errors:0 dropped:0 overruns:0 frame:0
282 TX packets:3286647 errors:1 dropped:0 overruns:1 carrier:0
283 collisions:0 txqueuelen:0
285 eth0Link encap:EthernetHwaddr 00:C0:F0:1F:37:B4
286 inet addr:XXX.XXX.XXX.YYY Bcast:XXX.XXX.XXX.255 Mask:255.255.252.0
287 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
288 RX packets:3573025 errors:0 dropped:0 overruns:0 frame:0
289 TX packets:1643167 errors:1 dropped:0 overruns:1 carrier:0
290 collisions:0 txqueuelen:100
291 Interrupt:10 Base address:0x1080
293 eth1Link encap:EthernetHwaddr 00:C0:F0:1F:37:B4
294 inet addr:XXX.XXX.XXX.YYY Bcast:XXX.XXX.XXX.255 Mask:255.255.252.0
295 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
296 RX packets:3651769 errors:0 dropped:0 overruns:0 frame:0
297 TX packets:1643480 errors:0 dropped:0 overruns:0 carrier:0
298 collisions:0 txqueuelen:100
299 Interrupt:9 Base address:0x1400</screen>
302 <section xml:id="dbdoclet.50438258_54769">
303 <title>Configuring Lustre with Bonding</title>
304 <para>Lustre uses the IP address of the bonded interfaces and requires no special configuration. It treats the bonded interface as a regular TCP/IP interface. If needed, specify bond0 using the Lustre networks parameter in /etc/modprobe.</para>
305 <screen>options lnet networks=tcp(bond0)</screen>
307 <section xml:id="dbdoclet.50438258_92244">
308 <title>Bonding References</title>
309 <para>We recommend the following bonding references:</para>
312 <para>In the Linux kernel source tree, see documentation/networking/bonding.txt</para>
315 <para><link xl:href="http://linux-ip.net/html/ether-bonding.html">http://linux-ip.net/html/ether-bonding.html</link></para>
318 <para><link xl:href="http://www.sourceforge.net/projects/bonding">http://www.sourceforge.net/projects/bonding</link></para>
321 <para>Linux Foundation bonding website: <link xl:href="http://www.linuxfoundation.org/collaborate/workgroups/networking/bonding">http://www.linuxfoundation.org/collaborate/workgroups/networking/bonding</link></para>
324 <para>This is the most extensive reference and we highly recommend it. This website includes explanations of more complicated setups, including the use of DHCP with bonding.</para>