Lines Matching +full:no +full:- +full:idle +full:- +full:on +full:- +full:init
1 .. SPDX-License-Identifier: GPL-2.0
11 Corrections, HA extensions: 2000/10/03-15:
13 - Willy Tarreau <willy at meta-x.org>
14 - Constantine Gavrilov <const-g at xpert.com>
15 - Chad N. Tindel <ctindel at ieee dot org>
16 - Janice Girouard <girouard at us dot ibm dot com>
17 - Jay Vosburgh <fubar at us dot ibm dot com>
22 - Mitch Williams <mitch.a.williams at intel.com>
35 the original tools from extreme-linux and beowulf sites will not work
77 8.3 Painfully Slow Or No Failed Link Detection By Miimon
119 -----------------------------------------------
123 (which is available on http://kernel.org). Most users "rolling their
135 ---------------------------
148 ``/etc/modprobe.d/*.conf`` configuration files, or in a distro-specific
151 Details on bonding support for sysfs is provided in the
156 configuring a bond, it is recommended "tail -f /var/log/messages" be
162 support at least miimon, so there is really no reason not to use it.
173 (active-backup, balance-alb and balance-tlb). Possible values
180 Note that this is only available through the sysfs interface. No module
184 active slave, or the empty string if there is no active slave or
190 is 1 - 65535. If the value is not specified, it takes 65535 as the
198 In an AD system, this specifies the mac-address for the actor in
200 address. If the all-zeroes MAC is specified, bonding will internally
202 local-admin bit set for this mac but driver does not enforce it. If
221 aggregator has no slaves.
230 - A slave is added to or removed from the bond
232 - Any slave's link state changes
234 - Any slave's 802.3ad association state changes
236 - The bond's administrative state changes to up
253 In an AD system, the port-key has three parts as shown below -
259 01-05 Speed
260 06-15 User-defined
264 from 0 - 1023. If not given, the system defaults to 0.
271 Specifies that duplicate frames (received on inactive ports) should be
274 Normally, bonding will drop duplicate frames (received on inactive
278 The default value is 0 (drop duplicate frames received on inactive
299 fashion, all replies from the ARP targets will be received on
314 default value is no IP addresses.
325 default value is no IPv6 addresses.
331 non-ARP traffic should be filtered (disregarded) for link
338 No validation or filtering is performed.
354 Filtering is applied to all slaves. No validation is
376 for backup slaves is on the broadcast ARP request sent out via the
382 The validation of ARP requests on backup slaves is mainly helping
421 This option affects only active-backup mode for slaves with
445 The default value is 2, and the allowable range is 1 - 255.
453 IEEE 802.1AX-2008 5.4.15 in addition to the existing coupled control
470 Specifies whether active-backup mode should set all slaves to
480 bonding to set all slaves of an active-backup bond to
497 The down side of this policy is that every device on
533 This option may be modified via sysfs only when no slaves are
545 on or 1
549 The default is on.
587 asserting carrier. It is similar to the Cisco EtherChannel min-links
589 must be up (link-up state) before marking the bond device as up
590 (carrier on). This is useful for situations where higher level services
604 balance-rr (round robin). Possible values are:
606 balance-rr or 0
608 Round-robin policy: Transmit packets in sequential
613 active-backup or 1
615 Active-backup policy: Only one slave in the bond is
618 externally visible on only one port (network adapter)
622 occurs in active-backup mode, bonding will issue one
623 or more gratuitous ARPs on the newly active slave.
634 balance-xor or 2
636 XOR policy: Transmit based on the selected transmit
647 Broadcast policy: transmits everything on all slave
662 regards to the packet mis-ordering requirements of
678 balance-tlb or 5
685 relative to the speed) on each slave.
687 In tlb_dynamic_lb=0 mode; the load balancing based on
700 balance-alb or 6
702 Adaptive load balancing: includes balance-tlb plus
707 the local system on their way out and overwrites the
730 and when an inactive slave is re-activated. The
763 failover event. As soon as the link is up on the new slave
764 (possibly immediately) a peer notification is sent on the
765 bonding device and each VLAN sub-device. This is repeated at
769 The valid range is 0 - 255; the default value is 1. These options
770 affect only the active-backup mode. These options were added for
783 The valid range is 0 - 65535; the default value is 1. This option
784 has effect only in balance-rr mode.
794 The valid range is 0 - 300000. The default value is 0, which means
803 for active-backup(1), balance-tlb (5) and balance-alb (6) mode.
813 off-line will alternate devices be used. This is useful when
817 The primary option is only valid for active-backup(1),
818 balance-tlb (5) and balance-alb (6) mode.
825 occurs. This option is designed to prevent flip-flopping between
847 If no slaves are active, the first slave to recover is
863 or alb mode. The value has no effect on any other modes.
866 slaves based on the load in that interval. This gives nice lb
867 characteristics but can cause packet reordering. If re-ordering is
868 a concern use this variable to disable flow shuffling and rely on
870 xmit-hash-policy can be used to select the appropriate hashing for
896 netif_carrier_ok() relies on the device driver to maintain its
903 "carrier on," so if a driver does not support netif_carrier,
915 balance-xor, 802.3ad, and tlb modes. Possible values are:
926 network peer on the same slave.
948 network peer on the same slave. For non-IP traffic,
981 information is omitted. For non-IP traffic, the
998 relies on skb_flow_dissect to obtain the header fields
1008 relies on skb_flow_dissect to obtain the header fields
1018 hash to load-balance traffic per-vlan, with failover
1021 use their own vlan, to give lacp-like functionality
1022 without requiring lacp-capable switching hardware.
1039 The valid range is 0 - 255; the default value is 1. A value of 0
1043 This option is useful for bonding modes balance-rr (0), active-backup
1044 (1), balance-tlb (5) and balance-alb (6), in which a failover can
1056 The valid range is 1 - 0x7fffffff; the default value is 1. This Option
1057 has effect only in balance-tlb and balance-alb modes.
1071 or partial support for bonding, then provide information on enabling
1076 initscripts or interfaces, or don't know if it's new enough, have no fear.
1085 $ rpm -qf /sbin/ifup
1100 ----------------------------------------
1111 slave devices. On SLES 9, this is most easily done by running the
1113 ifcfg-id file for each slave device. The simplest way to accomplish
1115 file ifcfg-id file created; see below for some issues with DHCP). The
1118 ifcfg-id-xx:xx:xx:xx:xx:xx
1123 Once the set of ifcfg-id-xx:xx:xx:xx:xx:xx files has been
1130 STARTMODE='on'
1131 USERCTL='no'
1133 _nm_name='bus-pci-0001:61:01.0'
1143 Once the ifcfg-id-xx:xx:xx:xx:xx:xx files have been modified,
1145 itself. This file is named ifcfg-bondX, where X is the number of the
1147 ifcfg-bond0, the second is ifcfg-bond1, and so on. The sysconfig
1151 The contents of the ifcfg-bondX file is as follows::
1161 BONDING_MODULE_OPTS="mode=active-backup miimon=100"
1163 BONDING_SLAVE1="bus-pci-0000:06:08.1"
1192 for the bonding mode, link monitoring, and so on here. Do not include
1202 (bus-pci-0000:06:08.1 in the example above) specify the physical
1212 # /etc/init.d/network restart
1220 devices (they do not show bonding interfaces on its list of network
1233 -------------------------------
1243 -----------------------------------------------
1247 bonding instance to have an appropriately configured ifcfg-bondX file
1251 ifcfg-bondX files.
1254 options in the ifcfg-bondX file, it is not necessary to add them to
1258 ------------------------------------------
1262 version 3 or later, Fedora, etc. On these systems, the network
1271 network-script file for all physical adapters that will be members of
1274 /etc/sysconfig/network-scripts
1276 The file name must be prefixed with "ifcfg-eth" and suffixed
1278 for eth0 would be named /etc/sysconfig/network-scripts/ifcfg-eth0.
1282 USERCTL=no
1289 must correspond with the name of the file, i.e., ifcfg-eth1 must have
1291 also depend on the final bonding interface name chosen for your bond.
1294 second is bond1, and so on.
1297 script will be /etc/sysconfig/network-scripts/ifcfg-bondX where X is
1298 the number of the bond. For bond0 the file is named "ifcfg-bond0",
1299 for bond1 it is named "ifcfg-bond1", and so on. Within that file,
1309 USERCTL=no
1316 and, indeed, preferable, to specify the bonding options in the ifcfg-bond0
1319 BONDING_OPTS="mode=active-backup arp_interval=60 arp_ip_target=192.168.1.254"
1342 options bond0 mode=balance-alb miimon=100
1347 Finally run "/etc/rc.d/init.d/network restart" as root. This
1352 ---------------------------------
1365 -------------------------------------------------
1369 specifying the appropriate BONDING_OPTS= in ifcfg-bondX where X is the
1377 -----------------------------------------------
1387 `ip link` commands to the system's global init script. The name of
1388 the global init script differs; for sysconfig, it is
1389 /etc/init.d/boot.local and for initscripts it is /etc/rc.d/rc.local.
1393 reboots, edit the appropriate file (/etc/init.d/boot.local or
1396 modprobe bonding mode=balance-alb miimon=100
1407 ifup and ifdown scripts on the bond devices. To reload the bonding
1410 # /etc/init.d/boot.local
1419 enabled without re-running the entire global init script.
1435 -----------------------------------------
1437 This section contains information on configuring multiple
1453 your distro uses these scripts, no special action is needed. See the
1464 options bond0 -o bond0 mode=balance-rr miimon=100
1467 options bond1 -o bond1 mode=balance-alb miimon=50
1470 named "bond0" and creates the bond0 device in balance-rr mode with an
1472 bond1 device in balance-alb mode with an miimon of 50.
1479 install bond1 /sbin/modprobe --ignore-install bonding -o bond1 \
1480 mode=balance-alb miimon=50
1486 to rename modules at load time (the "-o bond1" part). Attempts to pass
1488 This has been reported on some Fedora Core kernels, and has been seen on
1489 RHEL 4 as well. On kernels exhibiting this problem, it will be impossible
1494 ------------------------------------------
1499 allows for adding and removing bonds at runtime. Ifenslave is no
1514 -----------------------------
1521 # echo -bar > /sys/class/net/bonding_masters
1534 --------------------------
1546 # echo -eth0 > /sys/class/net/bond0/bonding/slaves
1555 # echo -eth0 > /sys/class/net/eth0/master/bonding/slaves
1560 -------------------------------
1564 The names of these files correspond directly with the command-
1573 To configure bond0 for balance-alb mode::
1577 - or -
1578 # echo balance-alb > /sys/class/net/bond0/bonding/mode
1584 To enable MII monitoring on bond0 with a 1 second interval::
1591 monitoring is enabled, and vice-versa.
1604 # echo -192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target
1617 ---------------------
1623 file (/etc/init.d/boot.local or /etc/rc.d/rc.local), and add the
1628 echo balance-alb > /sys/class/net/bond0/bonding/mode
1635 active-backup mode, using ARP monitoring, add the following lines to
1636 your init script::
1640 echo active-backup > /sys/class/net/bond1/bonding/mode
1648 -----------------------------------------
1654 The ifup and ifdown commands on Debian don't support bonding out of
1655 the box. The ifenslave-2.6 package should be installed to provide bonding
1656 support. Once installed, this package will provide ``bond-*`` options
1659 Note that ifenslave-2.6 package will load the bonding module and use
1663 ----------------------
1666 active-backup mode, with eth0 and eth1 as slaves::
1670 bond-slaves eth0 eth1
1671 bond-mode active-backup
1672 bond-miimon 100
1673 bond-primary eth0 eth1
1678 produce the same result on those systems::
1682 bond-slaves none
1683 bond-mode active-backup
1684 bond-miimon 100
1688 bond-master bond0
1689 bond-primary eth0 eth1
1693 bond-master bond0
1694 bond-primary eth0 eth1
1696 For a full list of ``bond-*`` supported options in /etc/network/interfaces and
1698 /usr/share/doc/ifenslave-2.6.
1701 ----------------------------------------------
1706 the selected bonding mode. On occasion however, it is helpful to direct certain
1707 classes of traffic to certain physical interfaces on output to implement
1718 tx_queues can be used to change this value. There is no sysfs parameter
1719 available as the allocation is done at module init time.
1724 Bonding Mode: fault-tolerance (active-backup)
1749 like the one above until proper priorities are set for all interfaces. On
1754 a multiqueue qdisc and filters to bias certain traffic to transmit on certain
1773 driver that is now present. This awareness allows tc filters to be placed on
1775 a pass-through for selecting output queues on the slave device rather than
1779 output slave selection was limited to round-robin and active-backup modes.
1782 ----------------------------------------------------------
1798 (a) ad_actor_system : You can set a random mac-address that can be used for
1800 Also it's preferable to set the local-admin bit. Following shell code
1801 generates a random mac-address as described above::
1813 is 65535, but system can take the value from 1 - 65535. Following shell
1819 (c) ad_user_port_key : Use the user portion of the port-key. The default
1820 keeps this empty. These are the upper 10 bits of the port-key and value
1821 ranges from 0 - 1023. Following shell code generates these 10 bits and
1832 -------------------------
1834 Each bonding device has a read-only file residing in the
1843 Bonding Mode: load balancing (round-robin)
1862 -------------------------
1867 contain information on which slaves are associated with which masters.
1905 The active-backup, balance-tlb and balance-alb modes do not
1916 The balance-rr, balance-xor and broadcast modes generally
1948 "un-accelerated" by the bonding driver so the VLAN tag sits in the
1951 VLAN interfaces *must* be added on top of a bonding interface
1955 would pick up the all-zeroes hardware address. Once the first slave
1960 are released from a bond that still has one or more VLAN interfaces on
1976 underlying device -- i.e. the bonding interface -- to promiscuous
1992 -------------------------
1995 queries to one or more designated peer systems on the network, and
1998 or more peers on the local network.
2001 ------------------------------------
2024 -------------------------
2033 then the MII monitor will rely on the driver for carrier state
2036 detect carrier loss on the device (e.g., when the cable is physically
2053 -------------------------
2071 case, another host on network 10 would use eth0 or eth1 before bond0).
2075 will be sent on one interface (bond0), but the corresponding reply
2076 will arrive on a different interface (eth0). This reply looks to ARP
2077 as an unsolicited ARP reply (because ARP matches replies on an
2088 ----------------------------
2090 On systems with network configuration scripts that do not
2099 options bond0 mode=some-mode miimon=50
2108 drivers are loaded next. Since no other drivers have been loaded,
2121 On systems utilizing modprobe an equivalent problem can occur.
2128 Full documentation on this can be found in the modprobe.d and modprobe
2131 8.3. Painfully Slow Or No Failed Link Detection By Miimon
2132 ---------------------------------------------------------
2155 carrier state. It has no way to determine the state of devices on or
2157 traffic while still maintaining carrier on.
2211 common to enable promiscuous mode on the device, so that all traffic
2217 For the balance-rr, balance-xor, broadcast, and 802.3ad modes,
2220 For the active-backup, balance-tlb and balance-alb modes, the
2223 For balance-tlb mode, the active slave is the slave currently
2226 For balance-alb mode, the active slave is the slave used as a
2227 "primary." This slave is used for mode-specific control traffic, for
2230 For the active-backup, balance-tlb and balance-alb modes, when
2245 --------------------------------------------------
2248 connected via multiple physical links, then there is no availability
2250 only one switch (or peer), so if it fails, there is no alternative
2256 for information on configuring bonding with one peer device.
2259 ----------------------------------------------------
2270 +-----+----+ +-----+----+
2272 | switch A +--------------------------+ switch B |
2274 +-----+----+ +-----++---+
2276 | +-------+ |
2277 +-------------+ host1 +---------------+
2278 eth0 +-------+ eth1
2282 the outside world ("port3" on each switch). There is no technical
2286 -------------------------------------------------------------
2288 In a topology such as the example above, the active-backup and
2290 availability; the other modes require all links to terminate on the
2293 active-backup:
2304 switches are not connected (no ISL), and the networks beyond
2306 necessary for some specific one-way traffic to reach both
2310 ----------------------------------------------------------------
2316 end, the MII monitor has no direct means to detect this. The ARP
2343 ------------------------------------------------------
2359 +----------+ +----------+
2361 | Host A +---------------------+ router +------------------->
2362 | +---------------------+ | Hosts B and C are out
2364 +----------+ +----------+
2373 and received via one other peer on the local network, the router.
2383 reach other stations on the same network. An example would be the
2386 +----------+ +----------+ +--------+
2387 | |eth0 port1| +-------+ Host B |
2388 | Host A +------------+ switch |port3 +--------+
2389 | +------------+ | +--------+
2390 | |eth1 port2| +------------------+ Host C |
2391 +----------+ +----------+port4 +--------+
2397 on the same local network (Hosts B and C in the above example).
2400 the bonded device will be to the same MAC level peer on the network
2414 -----------------------------------------------------------
2420 balance-rr:
2447 through the switch to a balance-rr bond will not utilize greater
2459 active-backup:
2461 the active-backup mode, as the inactive backup devices are all
2465 available bandwidth. On the plus side, active-backup mode
2470 balance-xor:
2475 configuration (as described above), with destinations all on
2480 As with balance-rr, the switch ports need to be configured for
2484 Like active-backup, there is not much advantage to this
2500 balance modes other than balance-rr, no single connection will
2508 traffic may also end up on a single device, but that is
2516 balance-tlb:
2517 The balance-tlb mode balances outgoing traffic by peer.
2523 manner (not a simple XOR as in balance-xor or 802.3ad mode),
2525 XOR to the same value) will not all "bunch up" on a single
2528 Unlike 802.3ad, interfaces may be of differing speeds, and no
2529 special switch configuration is required. On the down side,
2535 balance-alb:
2536 This mode is everything that balance-tlb is, and more.
2537 It has all of the features (and restrictions) of balance-tlb,
2547 ----------------------------------------------------
2556 -----------------------------------------------------
2562 +-----------+
2564 +-+---+---+-+
2566 +--------+ | +---------+
2568 +------+---+ +-----+----+ +-----+----+
2570 +------+---+ +-----+----+ +-----+----+
2572 +--------+ | +---------+
2574 +-+---+---+-+
2576 +-----------+
2582 cost effective than a single larger switch, e.g., on a network with 24
2591 -------------------------------------------------------------
2594 configurations of this type is balance-rr. Historically, in this
2599 packets has arrived). When employed in this fashion, the balance-rr
2604 ------------------------------------------------------
2617 -------------------------------------------
2636 Note that when a bonding interface has no active links, the
2643 cases with no connectivity, there is no additional penalty for
2652 --------------------------------
2660 idle for some period of time. This is most easily observed by issuing
2661 a "ping" to some other host on the network, and noticing that the
2664 For example, on a bond in active-backup mode with five slaves
2667 # ping -n 10.0.4.2
2683 the interfaces attached to the bond may occupy multiple ports on a
2689 switches exhibit this, and some do not. On switches that display this
2690 behavior, it can be induced by clearing the MAC forwarding table (on
2691 most Cisco switches, the privileged command "clear mac address-table
2698 bonding on specific hardware platforms, or for interfacing bonding
2702 --------------------
2706 On the JS20 blades, the bonding driver supports only
2707 balance-rr, active-backup, balance-tlb and balance-alb modes. This is
2712 --------------------------------
2715 integrated on the planar (that's "motherboard" in IBM-speak). In the
2718 An add-on Broadcom daughter card can be installed on a JS20 to provide
2727 Additional BladeCenter-specific networking information can be
2730 - "IBM eServer BladeCenter Networking Options"
2731 - "IBM eServer BladeCenter Layer 2-7 Network Switching"
2734 ------------------------------------
2759 -------------------------------
2761 The balance-rr mode requires the use of passthrough modules
2763 That switch must be configured for "etherchannel" or "trunking" on the
2764 appropriate ports, as is usual for balance-rr.
2766 The balance-alb and balance-tlb modes will function with
2773 The active-backup mode has no additional requirements.
2776 ----------------------
2783 ports and the devices on the JS20 system itself. The MII monitor is
2791 --------------
2799 It may be desirable to disable spanning tree on the switch
2801 avoid fail-over delay issues when using bonding.
2808 -------------------
2814 -----------------------------------------
2816 Any Ethernet type cards (you can even mix cards - a Intel
2821 slaves in active-backup mode.
2824 ----------------------------------------
2826 There is no limit.
2829 ----------------------------------------------
2836 ----------------------------------------
2839 disabled. The active-backup mode will fail over to a backup link, and
2842 manner is appropriate for the mode). See the sections on High
2850 monitors connectivity to another host on the local network.
2852 If no link monitoring is configured, the bonding driver will
2859 ----------------------------------------------
2861 Yes. See the section on High Availability for details.
2864 ---------------------------------------------
2868 In the basic balance modes (balance-rr and balance-xor), it
2873 The advanced balance modes (balance-tlb and balance-alb) do
2882 The active-backup mode should work with any Layer-II switch.
2885 ---------------------------------------------------------
2907 # ifconfig bond0 down ; modprobe -r bonding
2915 from the bond (``ifenslave -d bond0 eth0``). The bonding driver will
2923 version of the linux kernel, found on http://kernel.org
2929 on the main Linux network mailing list, hosted at vger.kernel.org. The list
2937 http://vger.kernel.org/vger-lists.html#netdev