blob: 080ad26690ae436a6981883cb48b0a624f63f478 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002 Linux Ethernet Bonding Driver HOWTO
3
Ben Hutchingsad246c92011-04-26 15:25:52 +00004 Latest update: 27 April 2011
Linus Torvalds1da177e2005-04-16 15:20:36 -07005
6Initial release : Thomas Davis <tadavis at lbl.gov>
7Corrections, HA extensions : 2000/10/03-15 :
8 - Willy Tarreau <willy at meta-x.org>
9 - Constantine Gavrilov <const-g at xpert.com>
10 - Chad N. Tindel <ctindel at ieee dot org>
11 - Janice Girouard <girouard at us dot ibm dot com>
12 - Jay Vosburgh <fubar at us dot ibm dot com>
13
14Reorganized and updated Feb 2005 by Jay Vosburgh
Auke Kok6224e012006-06-08 11:15:35 -070015Added Sysfs information: 2006/04/24
16 - Mitch Williams <mitch.a.williams at intel.com>
Linus Torvalds1da177e2005-04-16 15:20:36 -070017
Jay Vosburgh00354cf2005-07-21 12:18:02 -070018Introduction
19============
Linus Torvalds1da177e2005-04-16 15:20:36 -070020
Jay Vosburgh00354cf2005-07-21 12:18:02 -070021 The Linux bonding driver provides a method for aggregating
22multiple network interfaces into a single logical "bonded" interface.
23The behavior of the bonded interfaces depends upon the mode; generally
24speaking, modes provide either hot standby or load balancing services.
25Additionally, link integrity monitoring may be performed.
26
27 The bonding driver originally came from Donald Becker's
28beowulf patches for kernel 2.0. It has changed quite a bit since, and
29the original tools from extreme-linux and beowulf sites will not work
30with this version of the driver.
31
32 For new versions of the driver, updated userspace tools, and
33who to ask for help, please follow the links at the end of this file.
Linus Torvalds1da177e2005-04-16 15:20:36 -070034
35Table of Contents
36=================
37
381. Bonding Driver Installation
39
402. Bonding Driver Options
41
423. Configuring Bonding Devices
Auke Kok6224e012006-06-08 11:15:35 -0700433.1 Configuration with Sysconfig Support
443.1.1 Using DHCP with Sysconfig
453.1.2 Configuring Multiple Bonds with Sysconfig
463.2 Configuration with Initscripts Support
473.2.1 Using DHCP with Initscripts
483.2.2 Configuring Multiple Bonds with Initscripts
493.3 Configuring Bonding Manually with Ifenslave
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700503.3.1 Configuring Multiple Bonds Manually
Auke Kok6224e012006-06-08 11:15:35 -0700513.4 Configuring Bonding Manually via Sysfs
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000523.5 Configuration with Interfaces Support
533.6 Overriding Configuration for Special Cases
Linus Torvalds1da177e2005-04-16 15:20:36 -070054
Auke Kok6224e012006-06-08 11:15:35 -0700554. Querying Bonding Configuration
564.1 Bonding Configuration
574.2 Network Configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -070058
Auke Kok6224e012006-06-08 11:15:35 -0700595. Switch Configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -070060
Auke Kok6224e012006-06-08 11:15:35 -0700616. 802.1q VLAN Support
Linus Torvalds1da177e2005-04-16 15:20:36 -070062
Auke Kok6224e012006-06-08 11:15:35 -0700637. Link Monitoring
647.1 ARP Monitor Operation
657.2 Configuring Multiple ARP Targets
667.3 MII Monitor Operation
Linus Torvalds1da177e2005-04-16 15:20:36 -070067
Auke Kok6224e012006-06-08 11:15:35 -0700688. Potential Trouble Sources
698.1 Adventures in Routing
708.2 Ethernet Device Renaming
718.3 Painfully Slow Or No Failed Link Detection By Miimon
Linus Torvalds1da177e2005-04-16 15:20:36 -070072
Auke Kok6224e012006-06-08 11:15:35 -0700739. SNMP agents
Linus Torvalds1da177e2005-04-16 15:20:36 -070074
Auke Kok6224e012006-06-08 11:15:35 -07007510. Promiscuous mode
Linus Torvalds1da177e2005-04-16 15:20:36 -070076
Auke Kok6224e012006-06-08 11:15:35 -07007711. Configuring Bonding for High Availability
7811.1 High Availability in a Single Switch Topology
7911.2 High Availability in a Multiple Switch Topology
8011.2.1 HA Bonding Mode Selection for Multiple Switch Topology
8111.2.2 HA Link Monitoring for Multiple Switch Topology
Linus Torvalds1da177e2005-04-16 15:20:36 -070082
Auke Kok6224e012006-06-08 11:15:35 -07008312. Configuring Bonding for Maximum Throughput
8412.1 Maximum Throughput in a Single Switch Topology
8512.1.1 MT Bonding Mode Selection for Single Switch Topology
8612.1.2 MT Link Monitoring for Single Switch Topology
8712.2 Maximum Throughput in a Multiple Switch Topology
8812.2.1 MT Bonding Mode Selection for Multiple Switch Topology
8912.2.2 MT Link Monitoring for Multiple Switch Topology
Linus Torvalds1da177e2005-04-16 15:20:36 -070090
Auke Kok6224e012006-06-08 11:15:35 -07009113. Switch Behavior Issues
9213.1 Link Establishment and Failover Delays
9313.2 Duplicated Incoming Packets
Linus Torvalds1da177e2005-04-16 15:20:36 -070094
Auke Kok6224e012006-06-08 11:15:35 -07009514. Hardware Specific Considerations
9614.1 IBM BladeCenter
Jay Vosburgh00354cf2005-07-21 12:18:02 -070097
Auke Kok6224e012006-06-08 11:15:35 -07009815. Frequently Asked Questions
Jay Vosburgh00354cf2005-07-21 12:18:02 -070099
Auke Kok6224e012006-06-08 11:15:35 -070010016. Resources and Links
Linus Torvalds1da177e2005-04-16 15:20:36 -0700101
102
1031. Bonding Driver Installation
104==============================
105
106 Most popular distro kernels ship with the bonding driver
107already available as a module and the ifenslave user level control
108program installed and ready for use. If your distro does not, or you
109have need to compile bonding from source (e.g., configuring and
110installing a mainline kernel from kernel.org), you'll need to perform
111the following steps:
112
1131.1 Configure and build the kernel with bonding
114-----------------------------------------------
115
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700116 The current version of the bonding driver is available in the
Linus Torvalds1da177e2005-04-16 15:20:36 -0700117drivers/net/bonding subdirectory of the most recent kernel source
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700118(which is available on http://kernel.org). Most users "rolling their
119own" will want to use the most recent kernel from kernel.org.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700120
121 Configure kernel with "make menuconfig" (or "make xconfig" or
122"make config"), then select "Bonding driver support" in the "Network
123device support" section. It is recommended that you configure the
124driver as module since it is currently the only way to pass parameters
125to the driver or configure more than one bonding device.
126
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700127 Build and install the new kernel and modules, then continue
128below to install ifenslave.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700129
1301.2 Install ifenslave Control Utility
131-------------------------------------
132
133 The ifenslave user level control program is included in the
134kernel source tree, in the file Documentation/networking/ifenslave.c.
135It is generally recommended that you use the ifenslave that
136corresponds to the kernel that you are using (either from the same
137source tree or supplied with the distro), however, ifenslave
138executables from older kernels should function (but features newer
139than the ifenslave release are not supported). Running an ifenslave
140that is newer than the kernel is not supported, and may or may not
141work.
142
143 To install ifenslave, do the following:
144
145# gcc -Wall -O -I/usr/src/linux/include ifenslave.c -o ifenslave
146# cp ifenslave /sbin/ifenslave
147
148 If your kernel source is not in "/usr/src/linux," then replace
149"/usr/src/linux/include" in the above with the location of your kernel
150source include directory.
151
152 You may wish to back up any existing /sbin/ifenslave, or, for
153testing or informal use, tag the ifenslave to the kernel version
154(e.g., name the ifenslave executable /sbin/ifenslave-2.6.10).
155
156IMPORTANT NOTE:
157
158 If you omit the "-I" or specify an incorrect directory, you
159may end up with an ifenslave that is incompatible with the kernel
160you're trying to build it for. Some distros (e.g., Red Hat from 7.1
161onwards) do not have /usr/include/linux symbolically linked to the
162default kernel source include directory.
163
Auke Kok6224e012006-06-08 11:15:35 -0700164SECOND IMPORTANT NOTE:
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000165 If you plan to configure bonding using sysfs or using the
166/etc/network/interfaces file, you do not need to use ifenslave.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700167
1682. Bonding Driver Options
169=========================
170
Jay Vosburgh9a6c6862007-11-13 20:25:48 -0800171 Options for the bonding driver are supplied as parameters to the
172bonding module at load time, or are specified via sysfs.
173
174 Module options may be given as command line arguments to the
175insmod or modprobe command, but are usually specified in either the
176/etc/modules.conf or /etc/modprobe.conf configuration file, or in a
177distro-specific configuration file (some of which are detailed in the next
178section).
179
180 Details on bonding support for sysfs is provided in the
181"Configuring Bonding Manually via Sysfs" section, below.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700182
183 The available bonding driver parameters are listed below. If a
184parameter is not specified the default value is used. When initially
185configuring a bond, it is recommended "tail -f /var/log/messages" be
186run in a separate window to watch for bonding driver error messages.
187
188 It is critical that either the miimon or arp_interval and
189arp_ip_target parameters be specified, otherwise serious network
190degradation will occur during link failures. Very few devices do not
191support at least miimon, so there is really no reason not to use it.
192
193 Options with textual values will accept either the text name
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700194or, for backwards compatibility, the option value. E.g.,
195"mode=802.3ad" and "mode=4" set the same mode.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700196
197 The parameters are as follows:
198
Nicolas de Pesloüan1ba9ac72011-12-26 13:35:24 +0000199active_slave
200
201 Specifies the new active slave for modes that support it
202 (active-backup, balance-alb and balance-tlb). Possible values
203 are the name of any currently enslaved interface, or an empty
204 string. If a name is given, the slave and its link must be up in order
205 to be selected as the new active slave. If an empty string is
206 specified, the current active slave is cleared, and a new active
207 slave is selected automatically.
208
209 Note that this is only available through the sysfs interface. No module
210 parameter by this name exists.
211
212 The normal value of this option is the name of the currently
213 active slave, or the empty string if there is no active slave or
214 the current mode does not use an active slave.
215
Jay Vosburghfd989c82008-11-04 17:51:16 -0800216ad_select
217
218 Specifies the 802.3ad aggregation selection logic to use. The
219 possible values and their effects are:
220
221 stable or 0
222
223 The active aggregator is chosen by largest aggregate
224 bandwidth.
225
226 Reselection of the active aggregator occurs only when all
227 slaves of the active aggregator are down or the active
228 aggregator has no slaves.
229
230 This is the default value.
231
232 bandwidth or 1
233
234 The active aggregator is chosen by largest aggregate
235 bandwidth. Reselection occurs if:
236
237 - A slave is added to or removed from the bond
238
239 - Any slave's link state changes
240
241 - Any slave's 802.3ad association state changes
242
Matt LaPlante19f59462009-04-27 15:06:31 +0200243 - The bond's administrative state changes to up
Jay Vosburghfd989c82008-11-04 17:51:16 -0800244
245 count or 2
246
247 The active aggregator is chosen by the largest number of
248 ports (slaves). Reselection occurs as described under the
249 "bandwidth" setting, above.
250
251 The bandwidth and count selection policies permit failover of
252 802.3ad aggregations when partial failure of the active aggregator
253 occurs. This keeps the aggregator with the highest availability
254 (either in bandwidth or in number of ports) active at all times.
255
256 This option was added in bonding version 3.4.0.
257
Nicolas de Pesloüan025890b2011-08-06 07:06:39 +0000258all_slaves_active
259
260 Specifies that duplicate frames (received on inactive ports) should be
261 dropped (0) or delivered (1).
262
263 Normally, bonding will drop duplicate frames (received on inactive
264 ports), which is desirable for most users. But there are some times
265 it is nice to allow duplicate frames to be delivered.
266
267 The default value is 0 (drop duplicate frames received on inactive
268 ports).
269
Linus Torvalds1da177e2005-04-16 15:20:36 -0700270arp_interval
271
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700272 Specifies the ARP link monitoring frequency in milliseconds.
Jay Vosburghf5b2b962006-09-22 21:54:53 -0700273
274 The ARP monitor works by periodically checking the slave
275 devices to determine whether they have sent or received
276 traffic recently (the precise criteria depends upon the
277 bonding mode, and the state of the slave). Regular traffic is
278 generated via ARP probes issued for the addresses specified by
279 the arp_ip_target option.
280
281 This behavior can be modified by the arp_validate option,
282 below.
283
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700284 If ARP monitoring is used in an etherchannel compatible mode
285 (modes 0 and 2), the switch should be configured in a mode
286 that evenly distributes packets across all links. If the
287 switch is configured to distribute the packets in an XOR
Linus Torvalds1da177e2005-04-16 15:20:36 -0700288 fashion, all replies from the ARP targets will be received on
289 the same link which could cause the other team members to
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700290 fail. ARP monitoring should not be used in conjunction with
291 miimon. A value of 0 disables ARP monitoring. The default
Linus Torvalds1da177e2005-04-16 15:20:36 -0700292 value is 0.
293
294arp_ip_target
295
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700296 Specifies the IP addresses to use as ARP monitoring peers when
297 arp_interval is > 0. These are the targets of the ARP request
298 sent to determine the health of the link to the targets.
299 Specify these values in ddd.ddd.ddd.ddd format. Multiple IP
300 addresses must be separated by a comma. At least one IP
301 address must be given for ARP monitoring to function. The
302 maximum number of targets that can be specified is 16. The
303 default value is no IP addresses.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700304
Jay Vosburghf5b2b962006-09-22 21:54:53 -0700305arp_validate
306
307 Specifies whether or not ARP probes and replies should be
308 validated in the active-backup mode. This causes the ARP
309 monitor to examine the incoming ARP requests and replies, and
310 only consider a slave to be up if it is receiving the
311 appropriate ARP traffic.
312
313 Possible values are:
314
315 none or 0
316
317 No validation is performed. This is the default.
318
319 active or 1
320
321 Validation is performed only for the active slave.
322
323 backup or 2
324
325 Validation is performed only for backup slaves.
326
327 all or 3
328
329 Validation is performed for all slaves.
330
331 For the active slave, the validation checks ARP replies to
332 confirm that they were generated by an arp_ip_target. Since
333 backup slaves do not typically receive these replies, the
334 validation performed for backup slaves is on the ARP request
335 sent out via the active slave. It is possible that some
336 switch or network configurations may result in situations
337 wherein the backup slaves do not receive the ARP requests; in
338 such a situation, validation of backup slaves must be
339 disabled.
340
341 This option is useful in network configurations in which
342 multiple bonding hosts are concurrently issuing ARPs to one or
343 more targets beyond a common switch. Should the link between
344 the switch and target fail (but not the switch itself), the
345 probe traffic generated by the multiple bonding instances will
346 fool the standard ARP monitor into considering the links as
347 still up. Use of the arp_validate option can resolve this, as
348 the ARP monitor will only consider ARP requests and replies
349 associated with its own instance of bonding.
350
351 This option was added in bonding version 3.1.0.
352
Linus Torvalds1da177e2005-04-16 15:20:36 -0700353downdelay
354
355 Specifies the time, in milliseconds, to wait before disabling
356 a slave after a link failure has been detected. This option
357 is only valid for the miimon link monitor. The downdelay
358 value should be a multiple of the miimon value; if not, it
359 will be rounded down to the nearest multiple. The default
360 value is 0.
361
Jay Vosburghdd957c52007-10-09 19:57:24 -0700362fail_over_mac
363
364 Specifies whether active-backup mode should set all slaves to
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700365 the same MAC address at enslavement (the traditional
366 behavior), or, when enabled, perform special handling of the
367 bond's MAC address in accordance with the selected policy.
Jay Vosburghdd957c52007-10-09 19:57:24 -0700368
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700369 Possible values are:
Jay Vosburghdd957c52007-10-09 19:57:24 -0700370
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700371 none or 0
Jay Vosburghdd957c52007-10-09 19:57:24 -0700372
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700373 This setting disables fail_over_mac, and causes
374 bonding to set all slaves of an active-backup bond to
375 the same MAC address at enslavement time. This is the
376 default.
Jay Vosburghdd957c52007-10-09 19:57:24 -0700377
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700378 active or 1
Jay Vosburghdd957c52007-10-09 19:57:24 -0700379
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700380 The "active" fail_over_mac policy indicates that the
381 MAC address of the bond should always be the MAC
382 address of the currently active slave. The MAC
383 address of the slaves is not changed; instead, the MAC
384 address of the bond changes during a failover.
385
386 This policy is useful for devices that cannot ever
387 alter their MAC address, or for devices that refuse
388 incoming broadcasts with their own source MAC (which
389 interferes with the ARP monitor).
390
391 The down side of this policy is that every device on
392 the network must be updated via gratuitous ARP,
393 vs. just updating a switch or set of switches (which
394 often takes place for any traffic, not just ARP
395 traffic, if the switch snoops incoming traffic to
396 update its tables) for the traditional method. If the
397 gratuitous ARP is lost, communication may be
398 disrupted.
399
Lucas De Marchi25985ed2011-03-30 22:57:33 -0300400 When this policy is used in conjunction with the mii
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700401 monitor, devices which assert link up prior to being
402 able to actually transmit and receive are particularly
Matt LaPlante19f59462009-04-27 15:06:31 +0200403 susceptible to loss of the gratuitous ARP, and an
Jay Vosburgh3915c1e2008-05-17 21:10:14 -0700404 appropriate updelay setting may be required.
405
406 follow or 2
407
408 The "follow" fail_over_mac policy causes the MAC
409 address of the bond to be selected normally (normally
410 the MAC address of the first slave added to the bond).
411 However, the second and subsequent slaves are not set
412 to this MAC address while they are in a backup role; a
413 slave is programmed with the bond's MAC address at
414 failover time (and the formerly active slave receives
415 the newly active slave's MAC address).
416
417 This policy is useful for multiport devices that
418 either become confused or incur a performance penalty
419 when multiple ports are programmed with the same MAC
420 address.
421
422
423 The default policy is none, unless the first slave cannot
424 change its MAC address, in which case the active policy is
425 selected by default.
426
427 This option may be modified via sysfs only when no slaves are
428 present in the bond.
429
430 This option was added in bonding version 3.2.0. The "follow"
431 policy was added in bonding version 3.3.0.
Jay Vosburghdd957c52007-10-09 19:57:24 -0700432
Linus Torvalds1da177e2005-04-16 15:20:36 -0700433lacp_rate
434
435 Option specifying the rate in which we'll ask our link partner
436 to transmit LACPDU packets in 802.3ad mode. Possible values
437 are:
438
439 slow or 0
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700440 Request partner to transmit LACPDUs every 30 seconds
Linus Torvalds1da177e2005-04-16 15:20:36 -0700441
442 fast or 1
443 Request partner to transmit LACPDUs every 1 second
444
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700445 The default is slow.
446
Linus Torvalds1da177e2005-04-16 15:20:36 -0700447max_bonds
448
449 Specifies the number of bonding devices to create for this
450 instance of the bonding driver. E.g., if max_bonds is 3, and
451 the bonding driver is not already loaded, then bond0, bond1
Jay Vosburghb8a97872008-06-13 18:12:04 -0700452 and bond2 will be created. The default value is 1. Specifying
453 a value of 0 will load bonding, but will not create any devices.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700454
455miimon
456
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700457 Specifies the MII link monitoring frequency in milliseconds.
458 This determines how often the link state of each slave is
459 inspected for link failures. A value of zero disables MII
460 link monitoring. A value of 100 is a good starting point.
461 The use_carrier option, below, affects how the link state is
Linus Torvalds1da177e2005-04-16 15:20:36 -0700462 determined. See the High Availability section for additional
463 information. The default value is 0.
464
Nicolas de Pesloüan025890b2011-08-06 07:06:39 +0000465min_links
466
467 Specifies the minimum number of links that must be active before
468 asserting carrier. It is similar to the Cisco EtherChannel min-links
469 feature. This allows setting the minimum number of member ports that
470 must be up (link-up state) before marking the bond device as up
471 (carrier on). This is useful for situations where higher level services
472 such as clustering want to ensure a minimum number of low bandwidth
473 links are active before switchover. This option only affect 802.3ad
474 mode.
475
476 The default value is 0. This will cause carrier to be asserted (for
477 802.3ad mode) whenever there is an active aggregator, regardless of the
478 number of available links in that aggregator. Note that, because an
479 aggregator cannot be active without at least one available link,
480 setting this option to 0 or to 1 has the exact same effect.
481
Linus Torvalds1da177e2005-04-16 15:20:36 -0700482mode
483
484 Specifies one of the bonding policies. The default is
485 balance-rr (round robin). Possible values are:
486
487 balance-rr or 0
488
489 Round-robin policy: Transmit packets in sequential
490 order from the first available slave through the
491 last. This mode provides load balancing and fault
492 tolerance.
493
494 active-backup or 1
495
496 Active-backup policy: Only one slave in the bond is
497 active. A different slave becomes active if, and only
498 if, the active slave fails. The bond's MAC address is
499 externally visible on only one port (network adapter)
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700500 to avoid confusing the switch.
501
502 In bonding version 2.6.2 or later, when a failover
503 occurs in active-backup mode, bonding will issue one
504 or more gratuitous ARPs on the newly active slave.
Auke Kok6224e012006-06-08 11:15:35 -0700505 One gratuitous ARP is issued for the bonding master
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700506 interface and each VLAN interfaces configured above
507 it, provided that the interface has at least one IP
508 address configured. Gratuitous ARPs issued for VLAN
509 interfaces are tagged with the appropriate VLAN id.
510
511 This mode provides fault tolerance. The primary
512 option, documented below, affects the behavior of this
513 mode.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700514
515 balance-xor or 2
516
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700517 XOR policy: Transmit based on the selected transmit
518 hash policy. The default policy is a simple [(source
519 MAC address XOR'd with destination MAC address) modulo
520 slave count]. Alternate transmit policies may be
521 selected via the xmit_hash_policy option, described
522 below.
523
524 This mode provides load balancing and fault tolerance.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700525
526 broadcast or 3
527
528 Broadcast policy: transmits everything on all slave
529 interfaces. This mode provides fault tolerance.
530
531 802.3ad or 4
532
533 IEEE 802.3ad Dynamic link aggregation. Creates
534 aggregation groups that share the same speed and
535 duplex settings. Utilizes all slaves in the active
536 aggregator according to the 802.3ad specification.
537
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700538 Slave selection for outgoing traffic is done according
539 to the transmit hash policy, which may be changed from
540 the default simple XOR policy via the xmit_hash_policy
541 option, documented below. Note that not all transmit
542 policies may be 802.3ad compliant, particularly in
543 regards to the packet mis-ordering requirements of
544 section 43.2.4 of the 802.3ad standard. Differing
545 peer implementations will have varying tolerances for
546 noncompliance.
547
548 Prerequisites:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700549
550 1. Ethtool support in the base drivers for retrieving
551 the speed and duplex of each slave.
552
553 2. A switch that supports IEEE 802.3ad Dynamic link
554 aggregation.
555
556 Most switches will require some type of configuration
557 to enable 802.3ad mode.
558
559 balance-tlb or 5
560
561 Adaptive transmit load balancing: channel bonding that
562 does not require any special switch support. The
563 outgoing traffic is distributed according to the
564 current load (computed relative to the speed) on each
565 slave. Incoming traffic is received by the current
566 slave. If the receiving slave fails, another slave
567 takes over the MAC address of the failed receiving
568 slave.
569
570 Prerequisite:
571
572 Ethtool support in the base drivers for retrieving the
573 speed of each slave.
574
575 balance-alb or 6
576
577 Adaptive load balancing: includes balance-tlb plus
578 receive load balancing (rlb) for IPV4 traffic, and
579 does not require any special switch support. The
580 receive load balancing is achieved by ARP negotiation.
581 The bonding driver intercepts the ARP Replies sent by
582 the local system on their way out and overwrites the
583 source hardware address with the unique hardware
584 address of one of the slaves in the bond such that
585 different peers use different hardware addresses for
586 the server.
587
588 Receive traffic from connections created by the server
589 is also balanced. When the local system sends an ARP
590 Request the bonding driver copies and saves the peer's
591 IP information from the ARP packet. When the ARP
592 Reply arrives from the peer, its hardware address is
593 retrieved and the bonding driver initiates an ARP
594 reply to this peer assigning it to one of the slaves
595 in the bond. A problematic outcome of using ARP
596 negotiation for balancing is that each time that an
597 ARP request is broadcast it uses the hardware address
598 of the bond. Hence, peers learn the hardware address
599 of the bond and the balancing of receive traffic
600 collapses to the current slave. This is handled by
601 sending updates (ARP Replies) to all the peers with
602 their individually assigned hardware address such that
603 the traffic is redistributed. Receive traffic is also
604 redistributed when a new slave is added to the bond
605 and when an inactive slave is re-activated. The
606 receive load is distributed sequentially (round robin)
607 among the group of highest speed slaves in the bond.
608
609 When a link is reconnected or a new slave joins the
610 bond the receive traffic is redistributed among all
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700611 active slaves in the bond by initiating ARP Replies
Auke Kok6224e012006-06-08 11:15:35 -0700612 with the selected MAC address to each of the
Linus Torvalds1da177e2005-04-16 15:20:36 -0700613 clients. The updelay parameter (detailed below) must
614 be set to a value equal or greater than the switch's
615 forwarding delay so that the ARP Replies sent to the
616 peers will not be blocked by the switch.
617
618 Prerequisites:
619
620 1. Ethtool support in the base drivers for retrieving
621 the speed of each slave.
622
623 2. Base driver support for setting the hardware
624 address of a device while it is open. This is
625 required so that there will always be one slave in the
626 team using the bond hardware address (the
627 curr_active_slave) while having a unique hardware
628 address for each slave in the bond. If the
629 curr_active_slave fails its hardware address is
630 swapped with the new curr_active_slave that was
631 chosen.
632
Jay Vosburghb59f9f72008-06-13 18:12:03 -0700633num_grat_arp
Brian Haley305d5522008-11-04 17:51:14 -0800634num_unsol_na
635
Ben Hutchingsad246c92011-04-26 15:25:52 +0000636 Specify the number of peer notifications (gratuitous ARPs and
637 unsolicited IPv6 Neighbor Advertisements) to be issued after a
638 failover event. As soon as the link is up on the new slave
639 (possibly immediately) a peer notification is sent on the
640 bonding device and each VLAN sub-device. This is repeated at
641 each link monitor interval (arp_interval or miimon, whichever
642 is active) if the number is greater than 1.
Brian Haley305d5522008-11-04 17:51:14 -0800643
Ben Hutchingsad246c92011-04-26 15:25:52 +0000644 The valid range is 0 - 255; the default value is 1. These options
645 affect only the active-backup mode. These options were added for
646 bonding versions 3.3.0 and 3.4.0 respectively.
647
Jesper Juhl8fb4e132011-08-01 17:59:44 -0700648 From Linux 3.0 and bonding version 3.7.1, these notifications
Ben Hutchingsad246c92011-04-26 15:25:52 +0000649 are generated by the ipv4 and ipv6 code and the numbers of
650 repetitions cannot be set independently.
Brian Haley305d5522008-11-04 17:51:14 -0800651
Linus Torvalds1da177e2005-04-16 15:20:36 -0700652primary
653
654 A string (eth0, eth2, etc) specifying which slave is the
655 primary device. The specified device will always be the
656 active slave while it is available. Only when the primary is
657 off-line will alternate devices be used. This is useful when
658 one slave is preferred over another, e.g., when one slave has
659 higher throughput than another.
660
661 The primary option is only valid for active-backup mode.
662
Jiri Pirkoa5499522009-09-25 03:28:09 +0000663primary_reselect
664
665 Specifies the reselection policy for the primary slave. This
666 affects how the primary slave is chosen to become the active slave
667 when failure of the active slave or recovery of the primary slave
668 occurs. This option is designed to prevent flip-flopping between
669 the primary slave and other slaves. Possible values are:
670
671 always or 0 (default)
672
673 The primary slave becomes the active slave whenever it
674 comes back up.
675
676 better or 1
677
678 The primary slave becomes the active slave when it comes
679 back up, if the speed and duplex of the primary slave is
680 better than the speed and duplex of the current active
681 slave.
682
683 failure or 2
684
685 The primary slave becomes the active slave only if the
686 current active slave fails and the primary slave is up.
687
688 The primary_reselect setting is ignored in two cases:
689
690 If no slaves are active, the first slave to recover is
691 made the active slave.
692
693 When initially enslaved, the primary slave is always made
694 the active slave.
695
696 Changing the primary_reselect policy via sysfs will cause an
697 immediate selection of the best active slave according to the new
698 policy. This may or may not result in a change of the active
699 slave, depending upon the circumstances.
700
701 This option was added for bonding version 3.6.0.
702
Linus Torvalds1da177e2005-04-16 15:20:36 -0700703updelay
704
705 Specifies the time, in milliseconds, to wait before enabling a
706 slave after a link recovery has been detected. This option is
707 only valid for the miimon link monitor. The updelay value
708 should be a multiple of the miimon value; if not, it will be
709 rounded down to the nearest multiple. The default value is 0.
710
711use_carrier
712
713 Specifies whether or not miimon should use MII or ETHTOOL
714 ioctls vs. netif_carrier_ok() to determine the link
715 status. The MII or ETHTOOL ioctls are less efficient and
716 utilize a deprecated calling sequence within the kernel. The
717 netif_carrier_ok() relies on the device driver to maintain its
718 state with netif_carrier_on/off; at this writing, most, but
719 not all, device drivers support this facility.
720
721 If bonding insists that the link is up when it should not be,
722 it may be that your network device driver does not support
723 netif_carrier_on/off. The default state for netif_carrier is
724 "carrier on," so if a driver does not support netif_carrier,
725 it will appear as if the link is always up. In this case,
726 setting use_carrier to 0 will cause bonding to revert to the
727 MII / ETHTOOL ioctl method to determine the link state.
728
729 A value of 1 enables the use of netif_carrier_ok(), a value of
730 0 will use the deprecated MII / ETHTOOL ioctls. The default
731 value is 1.
732
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700733xmit_hash_policy
734
735 Selects the transmit hash policy to use for slave selection in
736 balance-xor and 802.3ad modes. Possible values are:
737
738 layer2
739
740 Uses XOR of hardware MAC addresses to generate the
741 hash. The formula is
742
743 (source MAC XOR destination MAC) modulo slave count
744
745 This algorithm will place all traffic to a particular
746 network peer on the same slave.
747
748 This algorithm is 802.3ad compliant.
749
Jay Vosburgh6f6652b2007-12-06 23:40:34 -0800750 layer2+3
751
752 This policy uses a combination of layer2 and layer3
753 protocol information to generate the hash.
754
755 Uses XOR of hardware MAC addresses and IP addresses to
756 generate the hash. The formula is
757
758 (((source IP XOR dest IP) AND 0xffff) XOR
759 ( source MAC XOR destination MAC ))
760 modulo slave count
761
762 This algorithm will place all traffic to a particular
763 network peer on the same slave. For non-IP traffic,
764 the formula is the same as for the layer2 transmit
765 hash policy.
766
767 This policy is intended to provide a more balanced
768 distribution of traffic than layer2 alone, especially
769 in environments where a layer3 gateway device is
770 required to reach most destinations.
771
Matt LaPlanted9195882008-07-25 19:45:33 -0700772 This algorithm is 802.3ad compliant.
Jay Vosburgh6f6652b2007-12-06 23:40:34 -0800773
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700774 layer3+4
775
776 This policy uses upper layer protocol information,
777 when available, to generate the hash. This allows for
778 traffic to a particular network peer to span multiple
779 slaves, although a single connection will not span
780 multiple slaves.
781
782 The formula for unfragmented TCP and UDP packets is
783
784 ((source port XOR dest port) XOR
785 ((source IP XOR dest IP) AND 0xffff)
786 modulo slave count
787
788 For fragmented TCP or UDP packets and all other IP
789 protocol traffic, the source and destination port
790 information is omitted. For non-IP traffic, the
791 formula is the same as for the layer2 transmit hash
792 policy.
793
794 This policy is intended to mimic the behavior of
795 certain switches, notably Cisco switches with PFC2 as
796 well as some Foundry and IBM products.
797
798 This algorithm is not fully 802.3ad compliant. A
799 single TCP or UDP conversation containing both
800 fragmented and unfragmented packets will see packets
801 striped across two interfaces. This may result in out
802 of order delivery. Most traffic types will not meet
803 this criteria, as TCP rarely fragments traffic, and
804 most UDP traffic is not involved in extended
805 conversations. Other implementations of 802.3ad may
806 or may not tolerate this noncompliance.
807
808 The default value is layer2. This option was added in bonding
Jay Vosburgh6f6652b2007-12-06 23:40:34 -0800809 version 2.6.3. In earlier versions of bonding, this parameter
810 does not exist, and the layer2 policy is the only policy. The
811 layer2+3 value was added for bonding version 3.2.2.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700812
Flavio Leitnerc2952c32010-10-05 14:23:59 +0000813resend_igmp
814
815 Specifies the number of IGMP membership reports to be issued after
816 a failover event. One membership report is issued immediately after
817 the failover, subsequent packets are sent in each 200ms interval.
818
Flavio Leitner94265cf2011-05-25 08:38:58 +0000819 The valid range is 0 - 255; the default value is 1. A value of 0
820 prevents the IGMP membership report from being issued in response
821 to the failover event.
822
823 This option is useful for bonding modes balance-rr (0), active-backup
824 (1), balance-tlb (5) and balance-alb (6), in which a failover can
825 switch the IGMP traffic from one slave to another. Therefore a fresh
826 IGMP report must be issued to cause the switch to forward the incoming
827 IGMP traffic over the newly selected slave.
828
829 This option was added for bonding version 3.7.0.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700830
8313. Configuring Bonding Devices
832==============================
833
Auke Kok6224e012006-06-08 11:15:35 -0700834 You can configure bonding using either your distro's network
835initialization scripts, or manually using either ifenslave or the
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000836sysfs interface. Distros generally use one of three packages for the
837network initialization scripts: initscripts, sysconfig or interfaces.
838Recent versions of these packages have support for bonding, while older
Auke Kok6224e012006-06-08 11:15:35 -0700839versions do not.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700840
841 We will first describe the options for configuring bonding for
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000842distros using versions of initscripts, sysconfig and interfaces with full
843or partial support for bonding, then provide information on enabling
Linus Torvalds1da177e2005-04-16 15:20:36 -0700844bonding without support from the network initialization scripts (i.e.,
845older versions of initscripts or sysconfig).
846
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000847 If you're unsure whether your distro uses sysconfig,
848initscripts or interfaces, or don't know if it's new enough, have no fear.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700849Determining this is fairly straightforward.
850
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +0000851 First, look for a file called interfaces in /etc/network directory.
852If this file is present in your system, then your system use interfaces. See
853Configuration with Interfaces Support.
854
855 Else, issue the command:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700856
857$ rpm -qf /sbin/ifup
858
859 It will respond with a line of text starting with either
860"initscripts" or "sysconfig," followed by some numbers. This is the
861package that provides your network initialization scripts.
862
863 Next, to determine if your installation supports bonding,
864issue the command:
865
866$ grep ifenslave /sbin/ifup
867
868 If this returns any matches, then your initscripts or
869sysconfig has support for bonding.
870
Auke Kok6224e012006-06-08 11:15:35 -07008713.1 Configuration with Sysconfig Support
Linus Torvalds1da177e2005-04-16 15:20:36 -0700872----------------------------------------
873
874 This section applies to distros using a version of sysconfig
875with bonding support, for example, SuSE Linux Enterprise Server 9.
876
877 SuSE SLES 9's networking configuration system does support
878bonding, however, at this writing, the YaST system configuration
Auke Kok6224e012006-06-08 11:15:35 -0700879front end does not provide any means to work with bonding devices.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700880Bonding devices can be managed by hand, however, as follows.
881
882 First, if they have not already been configured, configure the
883slave devices. On SLES 9, this is most easily done by running the
884yast2 sysconfig configuration utility. The goal is for to create an
885ifcfg-id file for each slave device. The simplest way to accomplish
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700886this is to configure the devices for DHCP (this is only to get the
887file ifcfg-id file created; see below for some issues with DHCP). The
888name of the configuration file for each device will be of the form:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700889
890ifcfg-id-xx:xx:xx:xx:xx:xx
891
892 Where the "xx" portion will be replaced with the digits from
893the device's permanent MAC address.
894
895 Once the set of ifcfg-id-xx:xx:xx:xx:xx:xx files has been
896created, it is necessary to edit the configuration files for the slave
897devices (the MAC addresses correspond to those of the slave devices).
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700898Before editing, the file will contain multiple lines, and will look
Linus Torvalds1da177e2005-04-16 15:20:36 -0700899something like this:
900
901BOOTPROTO='dhcp'
902STARTMODE='on'
903USERCTL='no'
904UNIQUE='XNzu.WeZGOGF+4wE'
905_nm_name='bus-pci-0001:61:01.0'
906
907 Change the BOOTPROTO and STARTMODE lines to the following:
908
909BOOTPROTO='none'
910STARTMODE='off'
911
912 Do not alter the UNIQUE or _nm_name lines. Remove any other
913lines (USERCTL, etc).
914
915 Once the ifcfg-id-xx:xx:xx:xx:xx:xx files have been modified,
916it's time to create the configuration file for the bonding device
917itself. This file is named ifcfg-bondX, where X is the number of the
918bonding device to create, starting at 0. The first such file is
919ifcfg-bond0, the second is ifcfg-bond1, and so on. The sysconfig
920network configuration system will correctly start multiple instances
921of bonding.
922
923 The contents of the ifcfg-bondX file is as follows:
924
925BOOTPROTO="static"
926BROADCAST="10.0.2.255"
927IPADDR="10.0.2.10"
928NETMASK="255.255.0.0"
929NETWORK="10.0.2.0"
930REMOTE_IPADDR=""
931STARTMODE="onboot"
932BONDING_MASTER="yes"
933BONDING_MODULE_OPTS="mode=active-backup miimon=100"
934BONDING_SLAVE0="eth0"
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700935BONDING_SLAVE1="bus-pci-0000:06:08.1"
Linus Torvalds1da177e2005-04-16 15:20:36 -0700936
937 Replace the sample BROADCAST, IPADDR, NETMASK and NETWORK
938values with the appropriate values for your network.
939
Linus Torvalds1da177e2005-04-16 15:20:36 -0700940 The STARTMODE specifies when the device is brought online.
941The possible values are:
942
943 onboot: The device is started at boot time. If you're not
944 sure, this is probably what you want.
945
946 manual: The device is started only when ifup is called
947 manually. Bonding devices may be configured this
948 way if you do not wish them to start automatically
949 at boot for some reason.
950
951 hotplug: The device is started by a hotplug event. This is not
952 a valid choice for a bonding device.
953
954 off or ignore: The device configuration is ignored.
955
956 The line BONDING_MASTER='yes' indicates that the device is a
957bonding master device. The only useful value is "yes."
958
959 The contents of BONDING_MODULE_OPTS are supplied to the
960instance of the bonding module for this device. Specify the options
961for the bonding mode, link monitoring, and so on here. Do not include
962the max_bonds bonding parameter; this will confuse the configuration
963system if you have multiple bonding devices.
964
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700965 Finally, supply one BONDING_SLAVEn="slave device" for each
966slave. where "n" is an increasing value, one for each slave. The
967"slave device" is either an interface name, e.g., "eth0", or a device
968specifier for the network device. The interface name is easier to
969find, but the ethN names are subject to change at boot time if, e.g.,
970a device early in the sequence has failed. The device specifiers
971(bus-pci-0000:06:08.1 in the example above) specify the physical
972network device, and will not change unless the device's bus location
973changes (for example, it is moved from one PCI slot to another). The
974example above uses one of each type for demonstration purposes; most
975configurations will choose one or the other for all slave devices.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700976
977 When all configuration files have been modified or created,
978networking must be restarted for the configuration changes to take
979effect. This can be accomplished via the following:
980
981# /etc/init.d/network restart
982
983 Note that the network control script (/sbin/ifdown) will
984remove the bonding module as part of the network shutdown processing,
985so it is not necessary to remove the module by hand if, e.g., the
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700986module parameters have changed.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700987
988 Also, at this writing, YaST/YaST2 will not manage bonding
989devices (they do not show bonding interfaces on its list of network
990devices). It is necessary to edit the configuration file by hand to
991change the bonding configuration.
992
993 Additional general options and details of the ifcfg file
994format can be found in an example ifcfg template file:
995
996/etc/sysconfig/network/ifcfg.template
997
998 Note that the template does not document the various BONDING_
999settings described above, but does describe many of the other options.
1000
Auke Kok6224e012006-06-08 11:15:35 -070010013.1.1 Using DHCP with Sysconfig
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001002-------------------------------
1003
1004 Under sysconfig, configuring a device with BOOTPROTO='dhcp'
1005will cause it to query DHCP for its IP address information. At this
1006writing, this does not function for bonding devices; the scripts
1007attempt to obtain the device address from DHCP prior to adding any of
1008the slave devices. Without active slaves, the DHCP requests are not
1009sent to the network.
1010
Auke Kok6224e012006-06-08 11:15:35 -070010113.1.2 Configuring Multiple Bonds with Sysconfig
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001012-----------------------------------------------
1013
1014 The sysconfig network initialization system is capable of
1015handling multiple bonding devices. All that is necessary is for each
1016bonding instance to have an appropriately configured ifcfg-bondX file
1017(as described above). Do not specify the "max_bonds" parameter to any
1018instance of bonding, as this will confuse sysconfig. If you require
1019multiple bonding devices with identical parameters, create multiple
1020ifcfg-bondX files.
1021
1022 Because the sysconfig scripts supply the bonding module
1023options in the ifcfg-bondX file, it is not necessary to add them to
1024the system /etc/modules.conf or /etc/modprobe.conf configuration file.
1025
Auke Kok6224e012006-06-08 11:15:35 -070010263.2 Configuration with Initscripts Support
Linus Torvalds1da177e2005-04-16 15:20:36 -07001027------------------------------------------
1028
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001029 This section applies to distros using a recent version of
1030initscripts with bonding support, for example, Red Hat Enterprise Linux
1031version 3 or later, Fedora, etc. On these systems, the network
1032initialization scripts have knowledge of bonding, and can be configured to
1033control bonding devices. Note that older versions of the initscripts
1034package have lower levels of support for bonding; this will be noted where
1035applicable.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001036
1037 These distros will not automatically load the network adapter
1038driver unless the ethX device is configured with an IP address.
1039Because of this constraint, users must manually configure a
1040network-script file for all physical adapters that will be members of
1041a bondX link. Network script files are located in the directory:
1042
1043/etc/sysconfig/network-scripts
1044
1045 The file name must be prefixed with "ifcfg-eth" and suffixed
1046with the adapter's physical adapter number. For example, the script
1047for eth0 would be named /etc/sysconfig/network-scripts/ifcfg-eth0.
1048Place the following text in the file:
1049
1050DEVICE=eth0
1051USERCTL=no
1052ONBOOT=yes
1053MASTER=bond0
1054SLAVE=yes
1055BOOTPROTO=none
1056
1057 The DEVICE= line will be different for every ethX device and
1058must correspond with the name of the file, i.e., ifcfg-eth1 must have
1059a device line of DEVICE=eth1. The setting of the MASTER= line will
1060also depend on the final bonding interface name chosen for your bond.
1061As with other network devices, these typically start at 0, and go up
1062one for each device, i.e., the first bonding instance is bond0, the
1063second is bond1, and so on.
1064
1065 Next, create a bond network script. The file name for this
1066script will be /etc/sysconfig/network-scripts/ifcfg-bondX where X is
1067the number of the bond. For bond0 the file is named "ifcfg-bond0",
1068for bond1 it is named "ifcfg-bond1", and so on. Within that file,
1069place the following text:
1070
1071DEVICE=bond0
1072IPADDR=192.168.1.1
1073NETMASK=255.255.255.0
1074NETWORK=192.168.1.0
1075BROADCAST=192.168.1.255
1076ONBOOT=yes
1077BOOTPROTO=none
1078USERCTL=no
1079
1080 Be sure to change the networking specific lines (IPADDR,
1081NETMASK, NETWORK and BROADCAST) to match your network configuration.
1082
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001083 For later versions of initscripts, such as that found with Fedora
Andy Gospodarek3f8b4b12008-10-22 11:19:48 +000010847 (or later) and Red Hat Enterprise Linux version 5 (or later), it is possible,
1085and, indeed, preferable, to specify the bonding options in the ifcfg-bond0
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001086file, e.g. a line of the format:
1087
Andy Gospodarek3f8b4b12008-10-22 11:19:48 +00001088BONDING_OPTS="mode=active-backup arp_interval=60 arp_ip_target=192.168.1.254"
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001089
1090 will configure the bond with the specified options. The options
1091specified in BONDING_OPTS are identical to the bonding module parameters
Andy Gospodarek3f8b4b12008-10-22 11:19:48 +00001092except for the arp_ip_target field when using versions of initscripts older
1093than and 8.57 (Fedora 8) and 8.45.19 (Red Hat Enterprise Linux 5.2). When
1094using older versions each target should be included as a separate option and
1095should be preceded by a '+' to indicate it should be added to the list of
1096queried targets, e.g.,
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001097
1098 arp_ip_target=+192.168.1.1 arp_ip_target=+192.168.1.2
1099
1100 is the proper syntax to specify multiple targets. When specifying
1101options via BONDING_OPTS, it is not necessary to edit /etc/modules.conf or
1102/etc/modprobe.conf.
1103
Andy Gospodarek3f8b4b12008-10-22 11:19:48 +00001104 For even older versions of initscripts that do not support
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001105BONDING_OPTS, it is necessary to edit /etc/modules.conf (or
1106/etc/modprobe.conf, depending upon your distro) to load the bonding module
1107with your desired options when the bond0 interface is brought up. The
1108following lines in /etc/modules.conf (or modprobe.conf) will load the
1109bonding module, and select its options:
Linus Torvalds1da177e2005-04-16 15:20:36 -07001110
1111alias bond0 bonding
1112options bond0 mode=balance-alb miimon=100
1113
1114 Replace the sample parameters with the appropriate set of
1115options for your configuration.
1116
1117 Finally run "/etc/rc.d/init.d/network restart" as root. This
1118will restart the networking subsystem and your bond link should be now
1119up and running.
1120
Auke Kok6224e012006-06-08 11:15:35 -070011213.2.1 Using DHCP with Initscripts
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001122---------------------------------
1123
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001124 Recent versions of initscripts (the versions supplied with Fedora
1125Core 3 and Red Hat Enterprise Linux 4, or later versions, are reported to
1126work) have support for assigning IP information to bonding devices via
1127DHCP.
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001128
1129 To configure bonding for DHCP, configure it as described
1130above, except replace the line "BOOTPROTO=none" with "BOOTPROTO=dhcp"
1131and add a line consisting of "TYPE=Bonding". Note that the TYPE value
1132is case sensitive.
1133
Auke Kok6224e012006-06-08 11:15:35 -070011343.2.2 Configuring Multiple Bonds with Initscripts
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001135-------------------------------------------------
1136
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001137 Initscripts packages that are included with Fedora 7 and Red Hat
1138Enterprise Linux 5 support multiple bonding interfaces by simply
1139specifying the appropriate BONDING_OPTS= in ifcfg-bondX where X is the
1140number of the bond. This support requires sysfs support in the kernel,
1141and a bonding driver of version 3.0.0 or later. Other configurations may
1142not support this method for specifying multiple bonding interfaces; for
1143those instances, see the "Configuring Multiple Bonds Manually" section,
1144below.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001145
Auke Kok6224e012006-06-08 11:15:35 -070011463.3 Configuring Bonding Manually with Ifenslave
1147-----------------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07001148
1149 This section applies to distros whose network initialization
1150scripts (the sysconfig or initscripts package) do not have specific
1151knowledge of bonding. One such distro is SuSE Linux Enterprise Server
1152version 8.
1153
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001154 The general method for these systems is to place the bonding
1155module parameters into /etc/modules.conf or /etc/modprobe.conf (as
1156appropriate for the installed distro), then add modprobe and/or
1157ifenslave commands to the system's global init script. The name of
1158the global init script differs; for sysconfig, it is
Linus Torvalds1da177e2005-04-16 15:20:36 -07001159/etc/init.d/boot.local and for initscripts it is /etc/rc.d/rc.local.
1160
1161 For example, if you wanted to make a simple bond of two e100
1162devices (presumed to be eth0 and eth1), and have it persist across
1163reboots, edit the appropriate file (/etc/init.d/boot.local or
1164/etc/rc.d/rc.local), and add the following:
1165
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001166modprobe bonding mode=balance-alb miimon=100
Linus Torvalds1da177e2005-04-16 15:20:36 -07001167modprobe e100
1168ifconfig bond0 192.168.1.1 netmask 255.255.255.0 up
1169ifenslave bond0 eth0
1170ifenslave bond0 eth1
1171
1172 Replace the example bonding module parameters and bond0
1173network configuration (IP address, netmask, etc) with the appropriate
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001174values for your configuration.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001175
1176 Unfortunately, this method will not provide support for the
1177ifup and ifdown scripts on the bond devices. To reload the bonding
1178configuration, it is necessary to run the initialization script, e.g.,
1179
1180# /etc/init.d/boot.local
1181
1182 or
1183
1184# /etc/rc.d/rc.local
1185
1186 It may be desirable in such a case to create a separate script
1187which only initializes the bonding configuration, then call that
1188separate script from within boot.local. This allows for bonding to be
1189enabled without re-running the entire global init script.
1190
1191 To shut down the bonding devices, it is necessary to first
1192mark the bonding device itself as being down, then remove the
1193appropriate device driver modules. For our example above, you can do
1194the following:
1195
1196# ifconfig bond0 down
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001197# rmmod bonding
Linus Torvalds1da177e2005-04-16 15:20:36 -07001198# rmmod e100
1199
1200 Again, for convenience, it may be desirable to create a script
1201with these commands.
1202
1203
Jay Vosburgh00354cf2005-07-21 12:18:02 -070012043.3.1 Configuring Multiple Bonds Manually
1205-----------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07001206
1207 This section contains information on configuring multiple
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001208bonding devices with differing options for those systems whose network
1209initialization scripts lack support for configuring multiple bonds.
1210
1211 If you require multiple bonding devices, but all with the same
1212options, you may wish to use the "max_bonds" module parameter,
1213documented above.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001214
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001215 To create multiple bonding devices with differing options, it is
1216preferrable to use bonding parameters exported by sysfs, documented in the
1217section below.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001218
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001219 For versions of bonding without sysfs support, the only means to
1220provide multiple instances of bonding with differing options is to load
1221the bonding driver multiple times. Note that current versions of the
1222sysconfig network initialization scripts handle this automatically; if
1223your distro uses these scripts, no special action is needed. See the
1224section Configuring Bonding Devices, above, if you're not sure about your
1225network initialization scripts.
1226
1227 To load multiple instances of the module, it is necessary to
1228specify a different name for each instance (the module loading system
1229requires that every loaded module, even multiple instances of the same
1230module, have a unique name). This is accomplished by supplying multiple
1231sets of bonding options in /etc/modprobe.conf, for example:
1232
1233alias bond0 bonding
1234options bond0 -o bond0 mode=balance-rr miimon=100
1235
1236alias bond1 bonding
1237options bond1 -o bond1 mode=balance-alb miimon=50
1238
1239 will load the bonding module two times. The first instance is
1240named "bond0" and creates the bond0 device in balance-rr mode with an
1241miimon of 100. The second instance is named "bond1" and creates the
1242bond1 device in balance-alb mode with an miimon of 50.
1243
1244 In some circumstances (typically with older distributions),
1245the above does not work, and the second bonding instance never sees
1246its options. In that case, the second options line can be substituted
1247as follows:
1248
1249install bond1 /sbin/modprobe --ignore-install bonding -o bond1 \
1250 mode=balance-alb miimon=50
1251
1252 This may be repeated any number of times, specifying a new and
1253unique name in place of bond1 for each subsequent instance.
1254
1255 It has been observed that some Red Hat supplied kernels are unable
1256to rename modules at load time (the "-o bond1" part). Attempts to pass
1257that option to modprobe will produce an "Operation not permitted" error.
1258This has been reported on some Fedora Core kernels, and has been seen on
1259RHEL 4 as well. On kernels exhibiting this problem, it will be impossible
1260to configure multiple bonds with differing parameters (as they are older
1261kernels, and also lack sysfs support).
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001262
Auke Kok6224e012006-06-08 11:15:35 -070012633.4 Configuring Bonding Manually via Sysfs
1264------------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07001265
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08001266 Starting with version 3.0.0, Channel Bonding may be configured
Auke Kok6224e012006-06-08 11:15:35 -07001267via the sysfs interface. This interface allows dynamic configuration
1268of all bonds in the system without unloading the module. It also
1269allows for adding and removing bonds at runtime. Ifenslave is no
1270longer required, though it is still supported.
1271
1272 Use of the sysfs interface allows you to use multiple bonds
1273with different configurations without having to reload the module.
1274It also allows you to use multiple, differently configured bonds when
1275bonding is compiled into the kernel.
1276
1277 You must have the sysfs filesystem mounted to configure
1278bonding this way. The examples in this document assume that you
1279are using the standard mount point for sysfs, e.g. /sys. If your
1280sysfs filesystem is mounted elsewhere, you will need to adjust the
1281example paths accordingly.
1282
1283Creating and Destroying Bonds
1284-----------------------------
1285To add a new bond foo:
1286# echo +foo > /sys/class/net/bonding_masters
1287
1288To remove an existing bond bar:
1289# echo -bar > /sys/class/net/bonding_masters
1290
1291To show all existing bonds:
1292# cat /sys/class/net/bonding_masters
1293
1294NOTE: due to 4K size limitation of sysfs files, this list may be
1295truncated if you have more than a few hundred bonds. This is unlikely
1296to occur under normal operating conditions.
1297
1298Adding and Removing Slaves
1299--------------------------
1300 Interfaces may be enslaved to a bond using the file
1301/sys/class/net/<bond>/bonding/slaves. The semantics for this file
1302are the same as for the bonding_masters file.
1303
1304To enslave interface eth0 to bond bond0:
1305# ifconfig bond0 up
1306# echo +eth0 > /sys/class/net/bond0/bonding/slaves
1307
1308To free slave eth0 from bond bond0:
1309# echo -eth0 > /sys/class/net/bond0/bonding/slaves
1310
Auke Kok6224e012006-06-08 11:15:35 -07001311 When an interface is enslaved to a bond, symlinks between the
1312two are created in the sysfs filesystem. In this case, you would get
1313/sys/class/net/bond0/slave_eth0 pointing to /sys/class/net/eth0, and
1314/sys/class/net/eth0/master pointing to /sys/class/net/bond0.
1315
1316 This means that you can tell quickly whether or not an
1317interface is enslaved by looking for the master symlink. Thus:
1318# echo -eth0 > /sys/class/net/eth0/master/bonding/slaves
1319will free eth0 from whatever bond it is enslaved to, regardless of
1320the name of the bond interface.
1321
1322Changing a Bond's Configuration
1323-------------------------------
1324 Each bond may be configured individually by manipulating the
1325files located in /sys/class/net/<bond name>/bonding
1326
1327 The names of these files correspond directly with the command-
Paolo Ornati670e9f32006-10-03 22:57:56 +02001328line parameters described elsewhere in this file, and, with the
Auke Kok6224e012006-06-08 11:15:35 -07001329exception of arp_ip_target, they accept the same values. To see the
1330current setting, simply cat the appropriate file.
1331
1332 A few examples will be given here; for specific usage
1333guidelines for each parameter, see the appropriate section in this
1334document.
1335
1336To configure bond0 for balance-alb mode:
1337# ifconfig bond0 down
1338# echo 6 > /sys/class/net/bond0/bonding/mode
1339 - or -
1340# echo balance-alb > /sys/class/net/bond0/bonding/mode
1341 NOTE: The bond interface must be down before the mode can be
1342changed.
1343
1344To enable MII monitoring on bond0 with a 1 second interval:
1345# echo 1000 > /sys/class/net/bond0/bonding/miimon
1346 NOTE: If ARP monitoring is enabled, it will disabled when MII
1347monitoring is enabled, and vice-versa.
1348
1349To add ARP targets:
1350# echo +192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target
1351# echo +192.168.0.101 > /sys/class/net/bond0/bonding/arp_ip_target
Brian Haley5a31bec2009-04-13 00:11:30 -07001352 NOTE: up to 16 target addresses may be specified.
Auke Kok6224e012006-06-08 11:15:35 -07001353
1354To remove an ARP target:
1355# echo -192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target
1356
1357Example Configuration
1358---------------------
1359 We begin with the same example that is shown in section 3.3,
1360executed with sysfs, and without using ifenslave.
1361
1362 To make a simple bond of two e100 devices (presumed to be eth0
1363and eth1), and have it persist across reboots, edit the appropriate
1364file (/etc/init.d/boot.local or /etc/rc.d/rc.local), and add the
1365following:
1366
1367modprobe bonding
1368modprobe e100
1369echo balance-alb > /sys/class/net/bond0/bonding/mode
1370ifconfig bond0 192.168.1.1 netmask 255.255.255.0 up
1371echo 100 > /sys/class/net/bond0/bonding/miimon
1372echo +eth0 > /sys/class/net/bond0/bonding/slaves
1373echo +eth1 > /sys/class/net/bond0/bonding/slaves
1374
1375 To add a second bond, with two e1000 interfaces in
1376active-backup mode, using ARP monitoring, add the following lines to
1377your init script:
1378
1379modprobe e1000
1380echo +bond1 > /sys/class/net/bonding_masters
1381echo active-backup > /sys/class/net/bond1/bonding/mode
1382ifconfig bond1 192.168.2.1 netmask 255.255.255.0 up
1383echo +192.168.2.100 /sys/class/net/bond1/bonding/arp_ip_target
1384echo 2000 > /sys/class/net/bond1/bonding/arp_interval
1385echo +eth2 > /sys/class/net/bond1/bonding/slaves
1386echo +eth3 > /sys/class/net/bond1/bonding/slaves
1387
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +000013883.5 Configuration with Interfaces Support
1389-----------------------------------------
1390
1391 This section applies to distros which use /etc/network/interfaces file
1392to describe network interface configuration, most notably Debian and it's
1393derivatives.
1394
1395 The ifup and ifdown commands on Debian don't support bonding out of
1396the box. The ifenslave-2.6 package should be installed to provide bonding
1397support. Once installed, this package will provide bond-* options to be used
1398into /etc/network/interfaces.
1399
1400 Note that ifenslave-2.6 package will load the bonding module and use
1401the ifenslave command when appropriate.
1402
1403Example Configurations
1404----------------------
1405
1406In /etc/network/interfaces, the following stanza will configure bond0, in
1407active-backup mode, with eth0 and eth1 as slaves.
1408
1409auto bond0
1410iface bond0 inet dhcp
1411 bond-slaves eth0 eth1
1412 bond-mode active-backup
1413 bond-miimon 100
1414 bond-primary eth0 eth1
1415
1416If the above configuration doesn't work, you might have a system using
1417upstart for system startup. This is most notably true for recent
1418Ubuntu versions. The following stanza in /etc/network/interfaces will
1419produce the same result on those systems.
1420
1421auto bond0
1422iface bond0 inet dhcp
1423 bond-slaves none
1424 bond-mode active-backup
1425 bond-miimon 100
1426
1427auto eth0
1428iface eth0 inet manual
1429 bond-master bond0
1430 bond-primary eth0 eth1
1431
1432auto eth1
1433iface eth1 inet manual
1434 bond-master bond0
1435 bond-primary eth0 eth1
1436
1437For a full list of bond-* supported options in /etc/network/interfaces and some
1438more advanced examples tailored to you particular distros, see the files in
1439/usr/share/doc/ifenslave-2.6.
1440
14413.6 Overriding Configuration for Special Cases
Andy Gospodarekbb1d9122010-06-02 08:40:18 +00001442----------------------------------------------
Nicolas de Pesloüande221bd2011-01-24 13:21:37 +00001443
Andy Gospodarekbb1d9122010-06-02 08:40:18 +00001444When using the bonding driver, the physical port which transmits a frame is
1445typically selected by the bonding driver, and is not relevant to the user or
1446system administrator. The output port is simply selected using the policies of
1447the selected bonding mode. On occasion however, it is helpful to direct certain
1448classes of traffic to certain physical interfaces on output to implement
1449slightly more complex policies. For example, to reach a web server over a
1450bonded interface in which eth0 connects to a private network, while eth1
1451connects via a public network, it may be desirous to bias the bond to send said
1452traffic over eth0 first, using eth1 only as a fall back, while all other traffic
1453can safely be sent over either interface. Such configurations may be achieved
1454using the traffic control utilities inherent in linux.
Auke Kok6224e012006-06-08 11:15:35 -07001455
Andy Gospodarekbb1d9122010-06-02 08:40:18 +00001456By default the bonding driver is multiqueue aware and 16 queues are created
1457when the driver initializes (see Documentation/networking/multiqueue.txt
1458for details). If more or less queues are desired the module parameter
1459tx_queues can be used to change this value. There is no sysfs parameter
1460available as the allocation is done at module init time.
1461
1462The output of the file /proc/net/bonding/bondX has changed so the output Queue
1463ID is now printed for each slave:
1464
1465Bonding Mode: fault-tolerance (active-backup)
1466Primary Slave: None
1467Currently Active Slave: eth0
1468MII Status: up
1469MII Polling Interval (ms): 0
1470Up Delay (ms): 0
1471Down Delay (ms): 0
1472
1473Slave Interface: eth0
1474MII Status: up
1475Link Failure Count: 0
1476Permanent HW addr: 00:1a:a0:12:8f:cb
1477Slave queue ID: 0
1478
1479Slave Interface: eth1
1480MII Status: up
1481Link Failure Count: 0
1482Permanent HW addr: 00:1a:a0:12:8f:cc
1483Slave queue ID: 2
1484
1485The queue_id for a slave can be set using the command:
1486
1487# echo "eth1:2" > /sys/class/net/bond0/bonding/queue_id
1488
1489Any interface that needs a queue_id set should set it with multiple calls
1490like the one above until proper priorities are set for all interfaces. On
1491distributions that allow configuration via initscripts, multiple 'queue_id'
1492arguments can be added to BONDING_OPTS to set all needed slave queues.
1493
1494These queue id's can be used in conjunction with the tc utility to configure
1495a multiqueue qdisc and filters to bias certain traffic to transmit on certain
1496slave devices. For instance, say we wanted, in the above configuration to
1497force all traffic bound to 192.168.1.100 to use eth1 in the bond as its output
1498device. The following commands would accomplish this:
1499
1500# tc qdisc add dev bond0 handle 1 root multiq
1501
1502# tc filter add dev bond0 protocol ip parent 1: prio 1 u32 match ip dst \
1503 192.168.1.100 action skbedit queue_mapping 2
1504
1505These commands tell the kernel to attach a multiqueue queue discipline to the
1506bond0 interface and filter traffic enqueued to it, such that packets with a dst
1507ip of 192.168.1.100 have their output queue mapping value overwritten to 2.
1508This value is then passed into the driver, causing the normal output path
1509selection policy to be overridden, selecting instead qid 2, which maps to eth1.
1510
1511Note that qid values begin at 1. Qid 0 is reserved to initiate to the driver
1512that normal output policy selection should take place. One benefit to simply
1513leaving the qid for a slave to 0 is the multiqueue awareness in the bonding
1514driver that is now present. This awareness allows tc filters to be placed on
1515slave devices as well as bond devices and the bonding driver will simply act as
1516a pass-through for selecting output queues on the slave device rather than
1517output port selection.
1518
1519This feature first appeared in bonding driver version 3.7.0 and support for
1520output slave selection was limited to round-robin and active-backup modes.
1521
15224 Querying Bonding Configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -07001523=================================
1524
Auke Kok6224e012006-06-08 11:15:35 -070015254.1 Bonding Configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -07001526-------------------------
1527
1528 Each bonding device has a read-only file residing in the
1529/proc/net/bonding directory. The file contents include information
1530about the bonding configuration, options and state of each slave.
1531
1532 For example, the contents of /proc/net/bonding/bond0 after the
1533driver is loaded with parameters of mode=0 and miimon=1000 is
1534generally as follows:
1535
1536 Ethernet Channel Bonding Driver: 2.6.1 (October 29, 2004)
1537 Bonding Mode: load balancing (round-robin)
1538 Currently Active Slave: eth0
1539 MII Status: up
1540 MII Polling Interval (ms): 1000
1541 Up Delay (ms): 0
1542 Down Delay (ms): 0
1543
1544 Slave Interface: eth1
1545 MII Status: up
1546 Link Failure Count: 1
1547
1548 Slave Interface: eth0
1549 MII Status: up
1550 Link Failure Count: 1
1551
1552 The precise format and contents will change depending upon the
1553bonding configuration, state, and version of the bonding driver.
1554
Auke Kok6224e012006-06-08 11:15:35 -070015554.2 Network configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -07001556-------------------------
1557
1558 The network configuration can be inspected using the ifconfig
1559command. Bonding devices will have the MASTER flag set; Bonding slave
1560devices will have the SLAVE flag set. The ifconfig output does not
1561contain information on which slaves are associated with which masters.
1562
1563 In the example below, the bond0 interface is the master
1564(MASTER) while eth0 and eth1 are slaves (SLAVE). Notice all slaves of
1565bond0 have the same MAC address (HWaddr) as bond0 for all modes except
1566TLB and ALB that require a unique MAC address for each slave.
1567
1568# /sbin/ifconfig
1569bond0 Link encap:Ethernet HWaddr 00:C0:F0:1F:37:B4
1570 inet addr:XXX.XXX.XXX.YYY Bcast:XXX.XXX.XXX.255 Mask:255.255.252.0
1571 UP BROADCAST RUNNING MASTER MULTICAST MTU:1500 Metric:1
1572 RX packets:7224794 errors:0 dropped:0 overruns:0 frame:0
1573 TX packets:3286647 errors:1 dropped:0 overruns:1 carrier:0
1574 collisions:0 txqueuelen:0
1575
1576eth0 Link encap:Ethernet HWaddr 00:C0:F0:1F:37:B4
Linus Torvalds1da177e2005-04-16 15:20:36 -07001577 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
1578 RX packets:3573025 errors:0 dropped:0 overruns:0 frame:0
1579 TX packets:1643167 errors:1 dropped:0 overruns:1 carrier:0
1580 collisions:0 txqueuelen:100
1581 Interrupt:10 Base address:0x1080
1582
1583eth1 Link encap:Ethernet HWaddr 00:C0:F0:1F:37:B4
Linus Torvalds1da177e2005-04-16 15:20:36 -07001584 UP BROADCAST RUNNING SLAVE MULTICAST MTU:1500 Metric:1
1585 RX packets:3651769 errors:0 dropped:0 overruns:0 frame:0
1586 TX packets:1643480 errors:0 dropped:0 overruns:0 carrier:0
1587 collisions:0 txqueuelen:100
1588 Interrupt:9 Base address:0x1400
1589
Auke Kok6224e012006-06-08 11:15:35 -070015905. Switch Configuration
Linus Torvalds1da177e2005-04-16 15:20:36 -07001591=======================
1592
1593 For this section, "switch" refers to whatever system the
1594bonded devices are directly connected to (i.e., where the other end of
1595the cable plugs into). This may be an actual dedicated switch device,
1596or it may be another regular system (e.g., another computer running
1597Linux),
1598
1599 The active-backup, balance-tlb and balance-alb modes do not
1600require any specific configuration of the switch.
1601
1602 The 802.3ad mode requires that the switch have the appropriate
1603ports configured as an 802.3ad aggregation. The precise method used
1604to configure this varies from switch to switch, but, for example, a
1605Cisco 3550 series switch requires that the appropriate ports first be
1606grouped together in a single etherchannel instance, then that
1607etherchannel is set to mode "lacp" to enable 802.3ad (instead of
1608standard EtherChannel).
1609
1610 The balance-rr, balance-xor and broadcast modes generally
1611require that the switch have the appropriate ports grouped together.
1612The nomenclature for such a group differs between switches, it may be
1613called an "etherchannel" (as in the Cisco example, above), a "trunk
1614group" or some other similar variation. For these modes, each switch
1615will also have its own configuration options for the switch's transmit
1616policy to the bond. Typical choices include XOR of either the MAC or
1617IP addresses. The transmit policy of the two peers does not need to
1618match. For these three modes, the bonding mode really selects a
1619transmit policy for an EtherChannel group; all three will interoperate
1620with another EtherChannel group.
1621
1622
Auke Kok6224e012006-06-08 11:15:35 -070016236. 802.1q VLAN Support
Linus Torvalds1da177e2005-04-16 15:20:36 -07001624======================
1625
1626 It is possible to configure VLAN devices over a bond interface
1627using the 8021q driver. However, only packets coming from the 8021q
1628driver and passing through bonding will be tagged by default. Self
1629generated packets, for example, bonding's learning packets or ARP
1630packets generated by either ALB mode or the ARP monitor mechanism, are
1631tagged internally by bonding itself. As a result, bonding must
1632"learn" the VLAN IDs configured above it, and use those IDs to tag
1633self generated packets.
1634
1635 For reasons of simplicity, and to support the use of adapters
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001636that can do VLAN hardware acceleration offloading, the bonding
1637interface declares itself as fully hardware offloading capable, it gets
Linus Torvalds1da177e2005-04-16 15:20:36 -07001638the add_vid/kill_vid notifications to gather the necessary
1639information, and it propagates those actions to the slaves. In case
1640of mixed adapter types, hardware accelerated tagged packets that
1641should go through an adapter that is not offloading capable are
1642"un-accelerated" by the bonding driver so the VLAN tag sits in the
1643regular location.
1644
1645 VLAN interfaces *must* be added on top of a bonding interface
1646only after enslaving at least one slave. The bonding interface has a
1647hardware address of 00:00:00:00:00:00 until the first slave is added.
1648If the VLAN interface is created prior to the first enslavement, it
1649would pick up the all-zeroes hardware address. Once the first slave
1650is attached to the bond, the bond device itself will pick up the
1651slave's hardware address, which is then available for the VLAN device.
1652
1653 Also, be aware that a similar problem can occur if all slaves
1654are released from a bond that still has one or more VLAN interfaces on
1655top of it. When a new slave is added, the bonding interface will
1656obtain its hardware address from the first slave, which might not
1657match the hardware address of the VLAN interfaces (which was
1658ultimately copied from an earlier slave).
1659
1660 There are two methods to insure that the VLAN device operates
1661with the correct hardware address if all slaves are removed from a
1662bond interface:
1663
1664 1. Remove all VLAN interfaces then recreate them
1665
1666 2. Set the bonding interface's hardware address so that it
1667matches the hardware address of the VLAN interfaces.
1668
1669 Note that changing a VLAN interface's HW address would set the
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001670underlying device -- i.e. the bonding interface -- to promiscuous
Linus Torvalds1da177e2005-04-16 15:20:36 -07001671mode, which might not be what you want.
1672
1673
Auke Kok6224e012006-06-08 11:15:35 -070016747. Link Monitoring
Linus Torvalds1da177e2005-04-16 15:20:36 -07001675==================
1676
1677 The bonding driver at present supports two schemes for
1678monitoring a slave device's link state: the ARP monitor and the MII
1679monitor.
1680
1681 At the present time, due to implementation restrictions in the
1682bonding driver itself, it is not possible to enable both ARP and MII
1683monitoring simultaneously.
1684
Auke Kok6224e012006-06-08 11:15:35 -070016857.1 ARP Monitor Operation
Linus Torvalds1da177e2005-04-16 15:20:36 -07001686-------------------------
1687
1688 The ARP monitor operates as its name suggests: it sends ARP
1689queries to one or more designated peer systems on the network, and
1690uses the response as an indication that the link is operating. This
1691gives some assurance that traffic is actually flowing to and from one
1692or more peers on the local network.
1693
1694 The ARP monitor relies on the device driver itself to verify
1695that traffic is flowing. In particular, the driver must keep up to
1696date the last receive time, dev->last_rx, and transmit start time,
1697dev->trans_start. If these are not updated by the driver, then the
1698ARP monitor will immediately fail any slaves using that driver, and
1699those slaves will stay down. If networking monitoring (tcpdump, etc)
1700shows the ARP requests and replies on the network, then it may be that
1701your device driver is not updating last_rx and trans_start.
1702
Auke Kok6224e012006-06-08 11:15:35 -070017037.2 Configuring Multiple ARP Targets
Linus Torvalds1da177e2005-04-16 15:20:36 -07001704------------------------------------
1705
1706 While ARP monitoring can be done with just one target, it can
1707be useful in a High Availability setup to have several targets to
1708monitor. In the case of just one target, the target itself may go
1709down or have a problem making it unresponsive to ARP requests. Having
1710an additional target (or several) increases the reliability of the ARP
1711monitoring.
1712
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001713 Multiple ARP targets must be separated by commas as follows:
Linus Torvalds1da177e2005-04-16 15:20:36 -07001714
1715# example options for ARP monitoring with three targets
1716alias bond0 bonding
1717options bond0 arp_interval=60 arp_ip_target=192.168.0.1,192.168.0.3,192.168.0.9
1718
1719 For just a single target the options would resemble:
1720
1721# example options for ARP monitoring with one target
1722alias bond0 bonding
1723options bond0 arp_interval=60 arp_ip_target=192.168.0.100
1724
1725
Auke Kok6224e012006-06-08 11:15:35 -070017267.3 MII Monitor Operation
Linus Torvalds1da177e2005-04-16 15:20:36 -07001727-------------------------
1728
1729 The MII monitor monitors only the carrier state of the local
1730network interface. It accomplishes this in one of three ways: by
1731depending upon the device driver to maintain its carrier state, by
1732querying the device's MII registers, or by making an ethtool query to
1733the device.
1734
1735 If the use_carrier module parameter is 1 (the default value),
1736then the MII monitor will rely on the driver for carrier state
1737information (via the netif_carrier subsystem). As explained in the
1738use_carrier parameter information, above, if the MII monitor fails to
1739detect carrier loss on the device (e.g., when the cable is physically
1740disconnected), it may be that the driver does not support
1741netif_carrier.
1742
1743 If use_carrier is 0, then the MII monitor will first query the
1744device's (via ioctl) MII registers and check the link state. If that
1745request fails (not just that it returns carrier down), then the MII
1746monitor will make an ethtool ETHOOL_GLINK request to attempt to obtain
1747the same information. If both methods fail (i.e., the driver either
1748does not support or had some error in processing both the MII register
1749and ethtool requests), then the MII monitor will assume the link is
1750up.
1751
Auke Kok6224e012006-06-08 11:15:35 -070017528. Potential Sources of Trouble
Linus Torvalds1da177e2005-04-16 15:20:36 -07001753===============================
1754
Auke Kok6224e012006-06-08 11:15:35 -070017558.1 Adventures in Routing
Linus Torvalds1da177e2005-04-16 15:20:36 -07001756-------------------------
1757
1758 When bonding is configured, it is important that the slave
Auke Kok6224e012006-06-08 11:15:35 -07001759devices not have routes that supersede routes of the master (or,
Linus Torvalds1da177e2005-04-16 15:20:36 -07001760generally, not have routes at all). For example, suppose the bonding
1761device bond0 has two slaves, eth0 and eth1, and the routing table is
1762as follows:
1763
1764Kernel IP routing table
1765Destination Gateway Genmask Flags MSS Window irtt Iface
176610.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 eth0
176710.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 eth1
176810.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 bond0
1769127.0.0.0 0.0.0.0 255.0.0.0 U 40 0 0 lo
1770
1771 This routing configuration will likely still update the
1772receive/transmit times in the driver (needed by the ARP monitor), but
1773may bypass the bonding driver (because outgoing traffic to, in this
1774case, another host on network 10 would use eth0 or eth1 before bond0).
1775
1776 The ARP monitor (and ARP itself) may become confused by this
1777configuration, because ARP requests (generated by the ARP monitor)
1778will be sent on one interface (bond0), but the corresponding reply
1779will arrive on a different interface (eth0). This reply looks to ARP
1780as an unsolicited ARP reply (because ARP matches replies on an
1781interface basis), and is discarded. The MII monitor is not affected
1782by the state of the routing table.
1783
1784 The solution here is simply to insure that slaves do not have
1785routes of their own, and if for some reason they must, those routes do
Auke Kok6224e012006-06-08 11:15:35 -07001786not supersede routes of their master. This should generally be the
Linus Torvalds1da177e2005-04-16 15:20:36 -07001787case, but unusual configurations or errant manual or automatic static
1788route additions may cause trouble.
1789
Auke Kok6224e012006-06-08 11:15:35 -070017908.2 Ethernet Device Renaming
Linus Torvalds1da177e2005-04-16 15:20:36 -07001791----------------------------
1792
1793 On systems with network configuration scripts that do not
1794associate physical devices directly with network interface names (so
1795that the same physical device always has the same "ethX" name), it may
1796be necessary to add some special logic to either /etc/modules.conf or
1797/etc/modprobe.conf (depending upon which is installed on the system).
1798
1799 For example, given a modules.conf containing the following:
1800
1801alias bond0 bonding
1802options bond0 mode=some-mode miimon=50
1803alias eth0 tg3
1804alias eth1 tg3
1805alias eth2 e1000
1806alias eth3 e1000
1807
1808 If neither eth0 and eth1 are slaves to bond0, then when the
1809bond0 interface comes up, the devices may end up reordered. This
1810happens because bonding is loaded first, then its slave device's
1811drivers are loaded next. Since no other drivers have been loaded,
1812when the e1000 driver loads, it will receive eth0 and eth1 for its
1813devices, but the bonding configuration tries to enslave eth2 and eth3
1814(which may later be assigned to the tg3 devices).
1815
1816 Adding the following:
1817
1818add above bonding e1000 tg3
1819
1820 causes modprobe to load e1000 then tg3, in that order, when
1821bonding is loaded. This command is fully documented in the
1822modules.conf manual page.
1823
1824 On systems utilizing modprobe.conf (or modprobe.conf.local),
1825an equivalent problem can occur. In this case, the following can be
1826added to modprobe.conf (or modprobe.conf.local, as appropriate), as
1827follows (all on one line; it has been split here for clarity):
1828
1829install bonding /sbin/modprobe tg3; /sbin/modprobe e1000;
1830 /sbin/modprobe --ignore-install bonding
1831
1832 This will, when loading the bonding module, rather than
1833performing the normal action, instead execute the provided command.
1834This command loads the device drivers in the order needed, then calls
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001835modprobe with --ignore-install to cause the normal action to then take
Linus Torvalds1da177e2005-04-16 15:20:36 -07001836place. Full documentation on this can be found in the modprobe.conf
1837and modprobe manual pages.
1838
Auke Kok6224e012006-06-08 11:15:35 -070018398.3. Painfully Slow Or No Failed Link Detection By Miimon
Linus Torvalds1da177e2005-04-16 15:20:36 -07001840---------------------------------------------------------
1841
1842 By default, bonding enables the use_carrier option, which
1843instructs bonding to trust the driver to maintain carrier state.
1844
1845 As discussed in the options section, above, some drivers do
1846not support the netif_carrier_on/_off link state tracking system.
1847With use_carrier enabled, bonding will always see these links as up,
1848regardless of their actual state.
1849
1850 Additionally, other drivers do support netif_carrier, but do
1851not maintain it in real time, e.g., only polling the link state at
1852some fixed interval. In this case, miimon will detect failures, but
1853only after some long period of time has expired. If it appears that
1854miimon is very slow in detecting link failures, try specifying
1855use_carrier=0 to see if that improves the failure detection time. If
1856it does, then it may be that the driver checks the carrier state at a
1857fixed interval, but does not cache the MII register values (so the
1858use_carrier=0 method of querying the registers directly works). If
1859use_carrier=0 does not improve the failover, then the driver may cache
1860the registers, or the problem may be elsewhere.
1861
1862 Also, remember that miimon only checks for the device's
1863carrier state. It has no way to determine the state of devices on or
1864beyond other ports of a switch, or if a switch is refusing to pass
1865traffic while still maintaining carrier on.
1866
Auke Kok6224e012006-06-08 11:15:35 -070018679. SNMP agents
Linus Torvalds1da177e2005-04-16 15:20:36 -07001868===============
1869
1870 If running SNMP agents, the bonding driver should be loaded
1871before any network drivers participating in a bond. This requirement
Tobias Klauserd533f672005-09-10 00:26:46 -07001872is due to the interface index (ipAdEntIfIndex) being associated to
Linus Torvalds1da177e2005-04-16 15:20:36 -07001873the first interface found with a given IP address. That is, there is
1874only one ipAdEntIfIndex for each IP address. For example, if eth0 and
1875eth1 are slaves of bond0 and the driver for eth0 is loaded before the
1876bonding driver, the interface for the IP address will be associated
1877with the eth0 interface. This configuration is shown below, the IP
1878address 192.168.1.1 has an interface index of 2 which indexes to eth0
1879in the ifDescr table (ifDescr.2).
1880
1881 interfaces.ifTable.ifEntry.ifDescr.1 = lo
1882 interfaces.ifTable.ifEntry.ifDescr.2 = eth0
1883 interfaces.ifTable.ifEntry.ifDescr.3 = eth1
1884 interfaces.ifTable.ifEntry.ifDescr.4 = eth2
1885 interfaces.ifTable.ifEntry.ifDescr.5 = eth3
1886 interfaces.ifTable.ifEntry.ifDescr.6 = bond0
1887 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.10.10.10.10 = 5
1888 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.192.168.1.1 = 2
1889 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.10.74.20.94 = 4
1890 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.127.0.0.1 = 1
1891
1892 This problem is avoided by loading the bonding driver before
1893any network drivers participating in a bond. Below is an example of
1894loading the bonding driver first, the IP address 192.168.1.1 is
1895correctly associated with ifDescr.2.
1896
1897 interfaces.ifTable.ifEntry.ifDescr.1 = lo
1898 interfaces.ifTable.ifEntry.ifDescr.2 = bond0
1899 interfaces.ifTable.ifEntry.ifDescr.3 = eth0
1900 interfaces.ifTable.ifEntry.ifDescr.4 = eth1
1901 interfaces.ifTable.ifEntry.ifDescr.5 = eth2
1902 interfaces.ifTable.ifEntry.ifDescr.6 = eth3
1903 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.10.10.10.10 = 6
1904 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.192.168.1.1 = 2
1905 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.10.74.20.94 = 5
1906 ip.ipAddrTable.ipAddrEntry.ipAdEntIfIndex.127.0.0.1 = 1
1907
1908 While some distributions may not report the interface name in
1909ifDescr, the association between the IP address and IfIndex remains
1910and SNMP functions such as Interface_Scan_Next will report that
1911association.
1912
Auke Kok6224e012006-06-08 11:15:35 -0700191310. Promiscuous mode
Linus Torvalds1da177e2005-04-16 15:20:36 -07001914====================
1915
1916 When running network monitoring tools, e.g., tcpdump, it is
1917common to enable promiscuous mode on the device, so that all traffic
1918is seen (instead of seeing only traffic destined for the local host).
1919The bonding driver handles promiscuous mode changes to the bonding
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001920master device (e.g., bond0), and propagates the setting to the slave
Linus Torvalds1da177e2005-04-16 15:20:36 -07001921devices.
1922
1923 For the balance-rr, balance-xor, broadcast, and 802.3ad modes,
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001924the promiscuous mode setting is propagated to all slaves.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001925
1926 For the active-backup, balance-tlb and balance-alb modes, the
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001927promiscuous mode setting is propagated only to the active slave.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001928
1929 For balance-tlb mode, the active slave is the slave currently
1930receiving inbound traffic.
1931
1932 For balance-alb mode, the active slave is the slave used as a
1933"primary." This slave is used for mode-specific control traffic, for
1934sending to peers that are unassigned or if the load is unbalanced.
1935
1936 For the active-backup, balance-tlb and balance-alb modes, when
1937the active slave changes (e.g., due to a link failure), the
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001938promiscuous setting will be propagated to the new active slave.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001939
Auke Kok6224e012006-06-08 11:15:35 -0700194011. Configuring Bonding for High Availability
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001941=============================================
Linus Torvalds1da177e2005-04-16 15:20:36 -07001942
1943 High Availability refers to configurations that provide
1944maximum network availability by having redundant or backup devices,
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001945links or switches between the host and the rest of the world. The
1946goal is to provide the maximum availability of network connectivity
1947(i.e., the network always works), even though other configurations
1948could provide higher throughput.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001949
Auke Kok6224e012006-06-08 11:15:35 -0700195011.1 High Availability in a Single Switch Topology
Linus Torvalds1da177e2005-04-16 15:20:36 -07001951--------------------------------------------------
1952
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001953 If two hosts (or a host and a single switch) are directly
1954connected via multiple physical links, then there is no availability
1955penalty to optimizing for maximum bandwidth. In this case, there is
1956only one switch (or peer), so if it fails, there is no alternative
1957access to fail over to. Additionally, the bonding load balance modes
1958support link monitoring of their members, so if individual links fail,
1959the load will be rebalanced across the remaining devices.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001960
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001961 See Section 13, "Configuring Bonding for Maximum Throughput"
1962for information on configuring bonding with one peer device.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001963
Auke Kok6224e012006-06-08 11:15:35 -0700196411.2 High Availability in a Multiple Switch Topology
Linus Torvalds1da177e2005-04-16 15:20:36 -07001965----------------------------------------------------
1966
1967 With multiple switches, the configuration of bonding and the
1968network changes dramatically. In multiple switch topologies, there is
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001969a trade off between network availability and usable bandwidth.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001970
1971 Below is a sample network, configured to maximize the
1972availability of the network:
1973
1974 | |
1975 |port3 port3|
1976 +-----+----+ +-----+----+
1977 | |port2 ISL port2| |
1978 | switch A +--------------------------+ switch B |
1979 | | | |
1980 +-----+----+ +-----++---+
1981 |port1 port1|
1982 | +-------+ |
1983 +-------------+ host1 +---------------+
1984 eth0 +-------+ eth1
1985
1986 In this configuration, there is a link between the two
1987switches (ISL, or inter switch link), and multiple ports connecting to
1988the outside world ("port3" on each switch). There is no technical
1989reason that this could not be extended to a third switch.
1990
Auke Kok6224e012006-06-08 11:15:35 -0700199111.2.1 HA Bonding Mode Selection for Multiple Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001992-------------------------------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07001993
Jay Vosburgh00354cf2005-07-21 12:18:02 -07001994 In a topology such as the example above, the active-backup and
1995broadcast modes are the only useful bonding modes when optimizing for
1996availability; the other modes require all links to terminate on the
1997same peer for them to behave rationally.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001998
1999active-backup: This is generally the preferred mode, particularly if
2000 the switches have an ISL and play together well. If the
2001 network configuration is such that one switch is specifically
2002 a backup switch (e.g., has lower capacity, higher cost, etc),
2003 then the primary option can be used to insure that the
2004 preferred link is always used when it is available.
2005
2006broadcast: This mode is really a special purpose mode, and is suitable
2007 only for very specific needs. For example, if the two
2008 switches are not connected (no ISL), and the networks beyond
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002009 them are totally independent. In this case, if it is
Linus Torvalds1da177e2005-04-16 15:20:36 -07002010 necessary for some specific one-way traffic to reach both
2011 independent networks, then the broadcast mode may be suitable.
2012
Auke Kok6224e012006-06-08 11:15:35 -0700201311.2.2 HA Link Monitoring Selection for Multiple Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002014----------------------------------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07002015
2016 The choice of link monitoring ultimately depends upon your
2017switch. If the switch can reliably fail ports in response to other
2018failures, then either the MII or ARP monitors should work. For
2019example, in the above example, if the "port3" link fails at the remote
2020end, the MII monitor has no direct means to detect this. The ARP
2021monitor could be configured with a target at the remote end of port3,
2022thus detecting that failure without switch support.
2023
2024 In general, however, in a multiple switch topology, the ARP
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002025monitor can provide a higher level of reliability in detecting end to
2026end connectivity failures (which may be caused by the failure of any
2027individual component to pass traffic for any reason). Additionally,
2028the ARP monitor should be configured with multiple targets (at least
2029one for each switch in the network). This will insure that,
Linus Torvalds1da177e2005-04-16 15:20:36 -07002030regardless of which switch is active, the ARP monitor has a suitable
2031target to query.
2032
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002033 Note, also, that of late many switches now support a functionality
2034generally referred to as "trunk failover." This is a feature of the
2035switch that causes the link state of a particular switch port to be set
2036down (or up) when the state of another switch port goes down (or up).
Matt LaPlante19f59462009-04-27 15:06:31 +02002037Its purpose is to propagate link failures from logically "exterior" ports
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002038to the logically "interior" ports that bonding is able to monitor via
2039miimon. Availability and configuration for trunk failover varies by
2040switch, but this can be a viable alternative to the ARP monitor when using
2041suitable switches.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002042
Auke Kok6224e012006-06-08 11:15:35 -0700204312. Configuring Bonding for Maximum Throughput
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002044==============================================
Linus Torvalds1da177e2005-04-16 15:20:36 -07002045
Auke Kok6224e012006-06-08 11:15:35 -0700204612.1 Maximizing Throughput in a Single Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002047------------------------------------------------------
2048
2049 In a single switch configuration, the best method to maximize
2050throughput depends upon the application and network environment. The
2051various load balancing modes each have strengths and weaknesses in
2052different environments, as detailed below.
2053
2054 For this discussion, we will break down the topologies into
2055two categories. Depending upon the destination of most traffic, we
2056categorize them into either "gatewayed" or "local" configurations.
2057
2058 In a gatewayed configuration, the "switch" is acting primarily
2059as a router, and the majority of traffic passes through this router to
2060other networks. An example would be the following:
2061
2062
2063 +----------+ +----------+
2064 | |eth0 port1| | to other networks
2065 | Host A +---------------------+ router +------------------->
2066 | +---------------------+ | Hosts B and C are out
2067 | |eth1 port2| | here somewhere
2068 +----------+ +----------+
2069
2070 The router may be a dedicated router device, or another host
2071acting as a gateway. For our discussion, the important point is that
2072the majority of traffic from Host A will pass through the router to
2073some other network before reaching its final destination.
2074
2075 In a gatewayed network configuration, although Host A may
2076communicate with many other systems, all of its traffic will be sent
2077and received via one other peer on the local network, the router.
2078
2079 Note that the case of two systems connected directly via
2080multiple physical links is, for purposes of configuring bonding, the
2081same as a gatewayed configuration. In that case, it happens that all
2082traffic is destined for the "gateway" itself, not some other network
2083beyond the gateway.
2084
2085 In a local configuration, the "switch" is acting primarily as
2086a switch, and the majority of traffic passes through this switch to
2087reach other stations on the same network. An example would be the
2088following:
2089
2090 +----------+ +----------+ +--------+
2091 | |eth0 port1| +-------+ Host B |
2092 | Host A +------------+ switch |port3 +--------+
2093 | +------------+ | +--------+
2094 | |eth1 port2| +------------------+ Host C |
2095 +----------+ +----------+port4 +--------+
2096
2097
2098 Again, the switch may be a dedicated switch device, or another
2099host acting as a gateway. For our discussion, the important point is
2100that the majority of traffic from Host A is destined for other hosts
2101on the same local network (Hosts B and C in the above example).
2102
2103 In summary, in a gatewayed configuration, traffic to and from
2104the bonded device will be to the same MAC level peer on the network
2105(the gateway itself, i.e., the router), regardless of its final
2106destination. In a local configuration, traffic flows directly to and
2107from the final destinations, thus, each destination (Host B, Host C)
2108will be addressed directly by their individual MAC addresses.
2109
2110 This distinction between a gatewayed and a local network
2111configuration is important because many of the load balancing modes
2112available use the MAC addresses of the local network source and
2113destination to make load balancing decisions. The behavior of each
2114mode is described below.
2115
2116
Auke Kok6224e012006-06-08 11:15:35 -0700211712.1.1 MT Bonding Mode Selection for Single Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002118-----------------------------------------------------------
2119
2120 This configuration is the easiest to set up and to understand,
2121although you will have to decide which bonding mode best suits your
2122needs. The trade offs for each mode are detailed below:
2123
2124balance-rr: This mode is the only mode that will permit a single
2125 TCP/IP connection to stripe traffic across multiple
2126 interfaces. It is therefore the only mode that will allow a
2127 single TCP/IP stream to utilize more than one interface's
2128 worth of throughput. This comes at a cost, however: the
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002129 striping generally results in peer systems receiving packets out
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002130 of order, causing TCP/IP's congestion control system to kick
2131 in, often by retransmitting segments.
2132
2133 It is possible to adjust TCP/IP's congestion limits by
2134 altering the net.ipv4.tcp_reordering sysctl parameter. The
2135 usual default value is 3, and the maximum useful value is 127.
2136 For a four interface balance-rr bond, expect that a single
2137 TCP/IP stream will utilize no more than approximately 2.3
2138 interface's worth of throughput, even after adjusting
2139 tcp_reordering.
2140
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002141 Note that the fraction of packets that will be delivered out of
2142 order is highly variable, and is unlikely to be zero. The level
2143 of reordering depends upon a variety of factors, including the
2144 networking interfaces, the switch, and the topology of the
2145 configuration. Speaking in general terms, higher speed network
2146 cards produce more reordering (due to factors such as packet
2147 coalescing), and a "many to many" topology will reorder at a
2148 higher rate than a "many slow to one fast" configuration.
2149
2150 Many switches do not support any modes that stripe traffic
2151 (instead choosing a port based upon IP or MAC level addresses);
2152 for those devices, traffic for a particular connection flowing
2153 through the switch to a balance-rr bond will not utilize greater
2154 than one interface's worth of bandwidth.
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002155
2156 If you are utilizing protocols other than TCP/IP, UDP for
2157 example, and your application can tolerate out of order
2158 delivery, then this mode can allow for single stream datagram
2159 performance that scales near linearly as interfaces are added
2160 to the bond.
2161
2162 This mode requires the switch to have the appropriate ports
2163 configured for "etherchannel" or "trunking."
2164
2165active-backup: There is not much advantage in this network topology to
2166 the active-backup mode, as the inactive backup devices are all
2167 connected to the same peer as the primary. In this case, a
2168 load balancing mode (with link monitoring) will provide the
2169 same level of network availability, but with increased
2170 available bandwidth. On the plus side, active-backup mode
2171 does not require any configuration of the switch, so it may
2172 have value if the hardware available does not support any of
2173 the load balance modes.
2174
2175balance-xor: This mode will limit traffic such that packets destined
2176 for specific peers will always be sent over the same
2177 interface. Since the destination is determined by the MAC
2178 addresses involved, this mode works best in a "local" network
2179 configuration (as described above), with destinations all on
2180 the same local network. This mode is likely to be suboptimal
2181 if all your traffic is passed through a single router (i.e., a
2182 "gatewayed" network configuration, as described above).
2183
2184 As with balance-rr, the switch ports need to be configured for
2185 "etherchannel" or "trunking."
2186
2187broadcast: Like active-backup, there is not much advantage to this
2188 mode in this type of network topology.
2189
2190802.3ad: This mode can be a good choice for this type of network
2191 topology. The 802.3ad mode is an IEEE standard, so all peers
2192 that implement 802.3ad should interoperate well. The 802.3ad
2193 protocol includes automatic configuration of the aggregates,
2194 so minimal manual configuration of the switch is needed
2195 (typically only to designate that some set of devices is
2196 available for 802.3ad). The 802.3ad standard also mandates
2197 that frames be delivered in order (within certain limits), so
2198 in general single connections will not see misordering of
2199 packets. The 802.3ad mode does have some drawbacks: the
2200 standard mandates that all devices in the aggregate operate at
2201 the same speed and duplex. Also, as with all bonding load
2202 balance modes other than balance-rr, no single connection will
2203 be able to utilize more than a single interface's worth of
2204 bandwidth.
2205
2206 Additionally, the linux bonding 802.3ad implementation
2207 distributes traffic by peer (using an XOR of MAC addresses),
2208 so in a "gatewayed" configuration, all outgoing traffic will
2209 generally use the same device. Incoming traffic may also end
2210 up on a single device, but that is dependent upon the
2211 balancing policy of the peer's 8023.ad implementation. In a
2212 "local" configuration, traffic will be distributed across the
2213 devices in the bond.
2214
2215 Finally, the 802.3ad mode mandates the use of the MII monitor,
2216 therefore, the ARP monitor is not available in this mode.
2217
2218balance-tlb: The balance-tlb mode balances outgoing traffic by peer.
2219 Since the balancing is done according to MAC address, in a
2220 "gatewayed" configuration (as described above), this mode will
2221 send all traffic across a single device. However, in a
2222 "local" network configuration, this mode balances multiple
2223 local network peers across devices in a vaguely intelligent
2224 manner (not a simple XOR as in balance-xor or 802.3ad mode),
2225 so that mathematically unlucky MAC addresses (i.e., ones that
2226 XOR to the same value) will not all "bunch up" on a single
2227 interface.
2228
2229 Unlike 802.3ad, interfaces may be of differing speeds, and no
2230 special switch configuration is required. On the down side,
2231 in this mode all incoming traffic arrives over a single
2232 interface, this mode requires certain ethtool support in the
2233 network device driver of the slave interfaces, and the ARP
2234 monitor is not available.
2235
2236balance-alb: This mode is everything that balance-tlb is, and more.
2237 It has all of the features (and restrictions) of balance-tlb,
2238 and will also balance incoming traffic from local network
2239 peers (as described in the Bonding Module Options section,
2240 above).
2241
2242 The only additional down side to this mode is that the network
2243 device driver must support changing the hardware address while
2244 the device is open.
2245
Auke Kok6224e012006-06-08 11:15:35 -0700224612.1.2 MT Link Monitoring for Single Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002247----------------------------------------------------
2248
2249 The choice of link monitoring may largely depend upon which
2250mode you choose to use. The more advanced load balancing modes do not
2251support the use of the ARP monitor, and are thus restricted to using
2252the MII monitor (which does not provide as high a level of end to end
2253assurance as the ARP monitor).
2254
Auke Kok6224e012006-06-08 11:15:35 -0700225512.2 Maximum Throughput in a Multiple Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002256-----------------------------------------------------
2257
2258 Multiple switches may be utilized to optimize for throughput
2259when they are configured in parallel as part of an isolated network
2260between two or more systems, for example:
2261
2262 +-----------+
2263 | Host A |
2264 +-+---+---+-+
2265 | | |
2266 +--------+ | +---------+
2267 | | |
2268 +------+---+ +-----+----+ +-----+----+
2269 | Switch A | | Switch B | | Switch C |
2270 +------+---+ +-----+----+ +-----+----+
2271 | | |
2272 +--------+ | +---------+
2273 | | |
2274 +-+---+---+-+
2275 | Host B |
2276 +-----------+
2277
2278 In this configuration, the switches are isolated from one
2279another. One reason to employ a topology such as this is for an
2280isolated network with many hosts (a cluster configured for high
2281performance, for example), using multiple smaller switches can be more
2282cost effective than a single larger switch, e.g., on a network with 24
2283hosts, three 24 port switches can be significantly less expensive than
2284a single 72 port switch.
2285
2286 If access beyond the network is required, an individual host
2287can be equipped with an additional network device connected to an
2288external network; this host then additionally acts as a gateway.
2289
Auke Kok6224e012006-06-08 11:15:35 -0700229012.2.1 MT Bonding Mode Selection for Multiple Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002291-------------------------------------------------------------
2292
2293 In actual practice, the bonding mode typically employed in
2294configurations of this type is balance-rr. Historically, in this
2295network configuration, the usual caveats about out of order packet
2296delivery are mitigated by the use of network adapters that do not do
2297any kind of packet coalescing (via the use of NAPI, or because the
2298device itself does not generate interrupts until some number of
2299packets has arrived). When employed in this fashion, the balance-rr
2300mode allows individual connections between two hosts to effectively
2301utilize greater than one interface's bandwidth.
2302
Auke Kok6224e012006-06-08 11:15:35 -0700230312.2.2 MT Link Monitoring for Multiple Switch Topology
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002304------------------------------------------------------
2305
2306 Again, in actual practice, the MII monitor is most often used
2307in this configuration, as performance is given preference over
2308availability. The ARP monitor will function in this topology, but its
2309advantages over the MII monitor are mitigated by the volume of probes
2310needed as the number of systems involved grows (remember that each
2311host in the network is configured with bonding).
2312
Auke Kok6224e012006-06-08 11:15:35 -0700231313. Switch Behavior Issues
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002314==========================
2315
Auke Kok6224e012006-06-08 11:15:35 -0700231613.1 Link Establishment and Failover Delays
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002317-------------------------------------------
2318
2319 Some switches exhibit undesirable behavior with regard to the
2320timing of link up and down reporting by the switch.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002321
2322 First, when a link comes up, some switches may indicate that
2323the link is up (carrier available), but not pass traffic over the
2324interface for some period of time. This delay is typically due to
2325some type of autonegotiation or routing protocol, but may also occur
2326during switch initialization (e.g., during recovery after a switch
2327failure). If you find this to be a problem, specify an appropriate
2328value to the updelay bonding module option to delay the use of the
2329relevant interface(s).
2330
2331 Second, some switches may "bounce" the link state one or more
2332times while a link is changing state. This occurs most commonly while
2333the switch is initializing. Again, an appropriate updelay value may
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002334help.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002335
2336 Note that when a bonding interface has no active links, the
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002337driver will immediately reuse the first link that goes up, even if the
2338updelay parameter has been specified (the updelay is ignored in this
2339case). If there are slave interfaces waiting for the updelay timeout
2340to expire, the interface that first went into that state will be
2341immediately reused. This reduces down time of the network if the
2342value of updelay has been overestimated, and since this occurs only in
2343cases with no connectivity, there is no additional penalty for
2344ignoring the updelay.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002345
2346 In addition to the concerns about switch timings, if your
2347switches take a long time to go into backup mode, it may be desirable
2348to not activate a backup interface immediately after a link goes down.
2349Failover may be delayed via the downdelay bonding module option.
2350
Auke Kok6224e012006-06-08 11:15:35 -0700235113.2 Duplicated Incoming Packets
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002352--------------------------------
2353
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002354 NOTE: Starting with version 3.0.2, the bonding driver has logic to
2355suppress duplicate packets, which should largely eliminate this problem.
2356The following description is kept for reference.
2357
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002358 It is not uncommon to observe a short burst of duplicated
2359traffic when the bonding device is first used, or after it has been
2360idle for some period of time. This is most easily observed by issuing
2361a "ping" to some other host on the network, and noticing that the
2362output from ping flags duplicates (typically one per slave).
2363
2364 For example, on a bond in active-backup mode with five slaves
2365all connected to one switch, the output may appear as follows:
2366
2367# ping -n 10.0.4.2
2368PING 10.0.4.2 (10.0.4.2) from 10.0.3.10 : 56(84) bytes of data.
236964 bytes from 10.0.4.2: icmp_seq=1 ttl=64 time=13.7 ms
237064 bytes from 10.0.4.2: icmp_seq=1 ttl=64 time=13.8 ms (DUP!)
237164 bytes from 10.0.4.2: icmp_seq=1 ttl=64 time=13.8 ms (DUP!)
237264 bytes from 10.0.4.2: icmp_seq=1 ttl=64 time=13.8 ms (DUP!)
237364 bytes from 10.0.4.2: icmp_seq=1 ttl=64 time=13.8 ms (DUP!)
237464 bytes from 10.0.4.2: icmp_seq=2 ttl=64 time=0.216 ms
237564 bytes from 10.0.4.2: icmp_seq=3 ttl=64 time=0.267 ms
237664 bytes from 10.0.4.2: icmp_seq=4 ttl=64 time=0.222 ms
2377
2378 This is not due to an error in the bonding driver, rather, it
2379is a side effect of how many switches update their MAC forwarding
2380tables. Initially, the switch does not associate the MAC address in
2381the packet with a particular switch port, and so it may send the
2382traffic to all ports until its MAC forwarding table is updated. Since
2383the interfaces attached to the bond may occupy multiple ports on a
2384single switch, when the switch (temporarily) floods the traffic to all
2385ports, the bond device receives multiple copies of the same packet
2386(one per slave device).
2387
2388 The duplicated packet behavior is switch dependent, some
2389switches exhibit this, and some do not. On switches that display this
2390behavior, it can be induced by clearing the MAC forwarding table (on
2391most Cisco switches, the privileged command "clear mac address-table
2392dynamic" will accomplish this).
2393
Auke Kok6224e012006-06-08 11:15:35 -0700239414. Hardware Specific Considerations
Linus Torvalds1da177e2005-04-16 15:20:36 -07002395====================================
2396
2397 This section contains additional information for configuring
2398bonding on specific hardware platforms, or for interfacing bonding
2399with particular switches or other devices.
2400
Auke Kok6224e012006-06-08 11:15:35 -0700240114.1 IBM BladeCenter
Linus Torvalds1da177e2005-04-16 15:20:36 -07002402--------------------
2403
2404 This applies to the JS20 and similar systems.
2405
2406 On the JS20 blades, the bonding driver supports only
2407balance-rr, active-backup, balance-tlb and balance-alb modes. This is
2408largely due to the network topology inside the BladeCenter, detailed
2409below.
2410
2411JS20 network adapter information
2412--------------------------------
2413
2414 All JS20s come with two Broadcom Gigabit Ethernet ports
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002415integrated on the planar (that's "motherboard" in IBM-speak). In the
2416BladeCenter chassis, the eth0 port of all JS20 blades is hard wired to
2417I/O Module #1; similarly, all eth1 ports are wired to I/O Module #2.
2418An add-on Broadcom daughter card can be installed on a JS20 to provide
2419two more Gigabit Ethernet ports. These ports, eth2 and eth3, are
2420wired to I/O Modules 3 and 4, respectively.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002421
2422 Each I/O Module may contain either a switch or a passthrough
2423module (which allows ports to be directly connected to an external
2424switch). Some bonding modes require a specific BladeCenter internal
2425network topology in order to function; these are detailed below.
2426
2427 Additional BladeCenter-specific networking information can be
2428found in two IBM Redbooks (www.ibm.com/redbooks):
2429
2430"IBM eServer BladeCenter Networking Options"
2431"IBM eServer BladeCenter Layer 2-7 Network Switching"
2432
2433BladeCenter networking configuration
2434------------------------------------
2435
2436 Because a BladeCenter can be configured in a very large number
2437of ways, this discussion will be confined to describing basic
2438configurations.
2439
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002440 Normally, Ethernet Switch Modules (ESMs) are used in I/O
Linus Torvalds1da177e2005-04-16 15:20:36 -07002441modules 1 and 2. In this configuration, the eth0 and eth1 ports of a
2442JS20 will be connected to different internal switches (in the
2443respective I/O modules).
2444
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002445 A passthrough module (OPM or CPM, optical or copper,
2446passthrough module) connects the I/O module directly to an external
2447switch. By using PMs in I/O module #1 and #2, the eth0 and eth1
2448interfaces of a JS20 can be redirected to the outside world and
2449connected to a common external switch.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002450
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002451 Depending upon the mix of ESMs and PMs, the network will
2452appear to bonding as either a single switch topology (all PMs) or as a
2453multiple switch topology (one or more ESMs, zero or more PMs). It is
2454also possible to connect ESMs together, resulting in a configuration
2455much like the example in "High Availability in a Multiple Switch
2456Topology," above.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002457
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002458Requirements for specific modes
2459-------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -07002460
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002461 The balance-rr mode requires the use of passthrough modules
2462for devices in the bond, all connected to an common external switch.
2463That switch must be configured for "etherchannel" or "trunking" on the
Linus Torvalds1da177e2005-04-16 15:20:36 -07002464appropriate ports, as is usual for balance-rr.
2465
2466 The balance-alb and balance-tlb modes will function with
2467either switch modules or passthrough modules (or a mix). The only
2468specific requirement for these modes is that all network interfaces
2469must be able to reach all destinations for traffic sent over the
2470bonding device (i.e., the network must converge at some point outside
2471the BladeCenter).
2472
2473 The active-backup mode has no additional requirements.
2474
2475Link monitoring issues
2476----------------------
2477
2478 When an Ethernet Switch Module is in place, only the ARP
2479monitor will reliably detect link loss to an external switch. This is
2480nothing unusual, but examination of the BladeCenter cabinet would
2481suggest that the "external" network ports are the ethernet ports for
2482the system, when it fact there is a switch between these "external"
2483ports and the devices on the JS20 system itself. The MII monitor is
2484only able to detect link failures between the ESM and the JS20 system.
2485
2486 When a passthrough module is in place, the MII monitor does
2487detect failures to the "external" port, which is then directly
2488connected to the JS20 system.
2489
2490Other concerns
2491--------------
2492
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002493 The Serial Over LAN (SoL) link is established over the primary
Linus Torvalds1da177e2005-04-16 15:20:36 -07002494ethernet (eth0) only, therefore, any loss of link to eth0 will result
2495in losing your SoL connection. It will not fail over with other
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002496network traffic, as the SoL system is beyond the control of the
2497bonding driver.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002498
2499 It may be desirable to disable spanning tree on the switch
2500(either the internal Ethernet Switch Module, or an external switch) to
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002501avoid fail-over delay issues when using bonding.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002502
2503
Auke Kok6224e012006-06-08 11:15:35 -0700250415. Frequently Asked Questions
Linus Torvalds1da177e2005-04-16 15:20:36 -07002505==============================
2506
25071. Is it SMP safe?
2508
2509 Yes. The old 2.0.xx channel bonding patch was not SMP safe.
2510The new driver was designed to be SMP safe from the start.
2511
25122. What type of cards will work with it?
2513
2514 Any Ethernet type cards (you can even mix cards - a Intel
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002515EtherExpress PRO/100 and a 3com 3c905b, for example). For most modes,
2516devices need not be of the same speed.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002517
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002518 Starting with version 3.2.1, bonding also supports Infiniband
2519slaves in active-backup mode.
2520
Linus Torvalds1da177e2005-04-16 15:20:36 -070025213. How many bonding devices can I have?
2522
2523 There is no limit.
2524
25254. How many slaves can a bonding device have?
2526
2527 This is limited only by the number of network interfaces Linux
2528supports and/or the number of network cards you can place in your
2529system.
2530
25315. What happens when a slave link dies?
2532
2533 If link monitoring is enabled, then the failing device will be
2534disabled. The active-backup mode will fail over to a backup link, and
2535other modes will ignore the failed link. The link will continue to be
2536monitored, and should it recover, it will rejoin the bond (in whatever
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002537manner is appropriate for the mode). See the sections on High
2538Availability and the documentation for each mode for additional
2539information.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002540
2541 Link monitoring can be enabled via either the miimon or
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002542arp_interval parameters (described in the module parameters section,
Linus Torvalds1da177e2005-04-16 15:20:36 -07002543above). In general, miimon monitors the carrier state as sensed by
2544the underlying network device, and the arp monitor (arp_interval)
2545monitors connectivity to another host on the local network.
2546
2547 If no link monitoring is configured, the bonding driver will
2548be unable to detect link failures, and will assume that all links are
2549always available. This will likely result in lost packets, and a
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002550resulting degradation of performance. The precise performance loss
Linus Torvalds1da177e2005-04-16 15:20:36 -07002551depends upon the bonding mode and network configuration.
2552
25536. Can bonding be used for High Availability?
2554
2555 Yes. See the section on High Availability for details.
2556
25577. Which switches/systems does it work with?
2558
2559 The full answer to this depends upon the desired mode.
2560
2561 In the basic balance modes (balance-rr and balance-xor), it
2562works with any system that supports etherchannel (also called
2563trunking). Most managed switches currently available have such
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002564support, and many unmanaged switches as well.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002565
2566 The advanced balance modes (balance-tlb and balance-alb) do
2567not have special switch requirements, but do need device drivers that
2568support specific features (described in the appropriate section under
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002569module parameters, above).
Linus Torvalds1da177e2005-04-16 15:20:36 -07002570
Auke Kok6224e012006-06-08 11:15:35 -07002571 In 802.3ad mode, it works with systems that support IEEE
Linus Torvalds1da177e2005-04-16 15:20:36 -07002572802.3ad Dynamic Link Aggregation. Most managed and many unmanaged
2573switches currently available support 802.3ad.
2574
2575 The active-backup mode should work with any Layer-II switch.
2576
25778. Where does a bonding device get its MAC address from?
2578
Jay Vosburgh9a6c6862007-11-13 20:25:48 -08002579 When using slave devices that have fixed MAC addresses, or when
2580the fail_over_mac option is enabled, the bonding device's MAC address is
2581the MAC address of the active slave.
2582
2583 For other configurations, if not explicitly configured (with
2584ifconfig or ip link), the MAC address of the bonding device is taken from
2585its first slave device. This MAC address is then passed to all following
2586slaves and remains persistent (even if the first slave is removed) until
2587the bonding device is brought down or reconfigured.
Linus Torvalds1da177e2005-04-16 15:20:36 -07002588
2589 If you wish to change the MAC address, you can set it with
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002590ifconfig or ip link:
Linus Torvalds1da177e2005-04-16 15:20:36 -07002591
2592# ifconfig bond0 hw ether 00:11:22:33:44:55
2593
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002594# ip link set bond0 address 66:77:88:99:aa:bb
2595
Linus Torvalds1da177e2005-04-16 15:20:36 -07002596 The MAC address can be also changed by bringing down/up the
2597device and then changing its slaves (or their order):
2598
2599# ifconfig bond0 down ; modprobe -r bonding
2600# ifconfig bond0 .... up
2601# ifenslave bond0 eth...
2602
2603 This method will automatically take the address from the next
2604slave that is added.
2605
2606 To restore your slaves' MAC addresses, you need to detach them
2607from the bond (`ifenslave -d bond0 eth0'). The bonding driver will
2608then restore the MAC addresses that the slaves had before they were
2609enslaved.
2610
Jay Vosburgh00354cf2005-07-21 12:18:02 -0700261116. Resources and Links
Linus Torvalds1da177e2005-04-16 15:20:36 -07002612=======================
2613
Nicolas de Pesloüana23c37f12011-03-13 10:34:22 +00002614 The latest version of the bonding driver can be found in the latest
Linus Torvalds1da177e2005-04-16 15:20:36 -07002615version of the linux kernel, found on http://kernel.org
2616
Nicolas de Pesloüana23c37f12011-03-13 10:34:22 +00002617 The latest version of this document can be found in the latest kernel
2618source (named Documentation/networking/bonding.txt).
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002619
Nicolas de Pesloüana23c37f12011-03-13 10:34:22 +00002620 Discussions regarding the usage of the bonding driver take place on the
2621bonding-devel mailing list, hosted at sourceforge.net. If you have questions or
2622problems, post them to the list. The list address is:
Linus Torvalds1da177e2005-04-16 15:20:36 -07002623
2624bonding-devel@lists.sourceforge.net
2625
Jay Vosburgh00354cf2005-07-21 12:18:02 -07002626 The administrative interface (to subscribe or unsubscribe) can
2627be found at:
2628
Linus Torvalds1da177e2005-04-16 15:20:36 -07002629https://lists.sourceforge.net/lists/listinfo/bonding-devel
2630
Nicolas de Pesloüana23c37f12011-03-13 10:34:22 +00002631 Discussions regarding the developpement of the bonding driver take place
2632on the main Linux network mailing list, hosted at vger.kernel.org. The list
2633address is:
2634
2635netdev@vger.kernel.org
2636
2637 The administrative interface (to subscribe or unsubscribe) can
2638be found at:
2639
2640http://vger.kernel.org/vger-lists.html#netdev
2641
Linus Torvalds1da177e2005-04-16 15:20:36 -07002642Donald Becker's Ethernet Drivers and diag programs may be found at :
Justin P. Mattock0ea6e612010-07-23 20:51:24 -07002643 - http://web.archive.org/web/*/http://www.scyld.com/network/
Linus Torvalds1da177e2005-04-16 15:20:36 -07002644
2645You will also find a lot of information regarding Ethernet, NWay, MII,
2646etc. at www.scyld.com.
2647
2648-- END --