1.. BSD LICENSE 2 Copyright(c) 2010-2014 Intel Corporation. All rights reserved. 3 All rights reserved. 4 5 Redistribution and use in source and binary forms, with or without 6 modification, are permitted provided that the following conditions 7 are met: 8 9 * Redistributions of source code must retain the above copyright 10 notice, this list of conditions and the following disclaimer. 11 * Redistributions in binary form must reproduce the above copyright 12 notice, this list of conditions and the following disclaimer in 13 the documentation and/or other materials provided with the 14 distribution. 15 * Neither the name of Intel Corporation nor the names of its 16 contributors may be used to endorse or promote products derived 17 from this software without specific prior written permission. 18 19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 30 31Link Bonding Poll Mode Driver Library 32===================================== 33 34In addition to Poll Mode Drivers (PMDs) for physical and virtual hardware, 35Intel® DPDK also includes a pure-software library that 36allows physical PMD's to be bonded together to create a single logical PMD. 37 38|link_bonding| 39 40The Link Bonding PMD library(librte_pmd_bond) supports bonding of groups of physical ports of the same speed (1GbE, 10GbE and 40GbE) and 41duplex to provide similar the capabilities to that found in Linux bonding driver to allow the aggregation of multiple (slave) NICs 42into a single logical interface between a server and a switch. 43The new bonded PMD will then process these interfaces based on the mode of operation specified to provide support for features 44such as redundant links, fault tolerance and/or load balancing. 45 46The librte_pmd_bond library exports a C API which provides an API for the creation of bonded devices 47as well as the configuration and management of the bonded device and its slave devices. 48 49.. note:: 50 51 The Link Bonding PMD Library is enabled by default in the build configuration files, 52 the library can be disabled by setting CONFIG_RTE_LIBRTE_PMD_BOND=n and recompiling the Intel® DPDK. 53 54Link Bonding Modes Overview 55--------------------------- 56 57Currently the Link Bonding PMD library supports 4 modes of operation: 58 59* **Round-Robin (Mode 0):** 60 This mode provides load balancing and fault tolerance by transmission of packets 61 in sequential order from the first available slave device through the last. 62 Packets are bulk dequeued from devices then serviced in round-robin manner. 63 64* **Active Backup (Mode 1):** 65 In this mode only one slave in the bond is active at any time, a different slave becomes active if, 66 and only if, the primary active slave fails, 67 thereby providing fault tolerance to slave failure. 68 The single logical bonded interface's MAC address is externally visible on only one NIC (port) 69 to avoid confusing the network switch. 70 71* **Balance XOR (Mode 2):** 72 This mode provides load balancing based on transmit packets based on the selected XOR transmission policy and fault tolerance. 73 The default policy (layer2) uses a simple XOR calculation on the packet source / destination MAC address to select the slave to transmit on. 74 Alternate transmission policies supported are layer 2+3, this uses the IP source and destination addresses in the calculation of the slave port and 75 the final supported policy is layer 3+4, this uses IP source and destination addresses as well as the UDP source and destination port. 76 77* **Broadcast (Mode 3):** 78 This mode provides fault tolerance by transmission of packets on all slave ports. 79 80Implementation Details 81---------------------- 82 83The librte_pmd_bond onded device are compatible with the Ethernet device API exported by the Ethernet PMDs described in the *Intel® DPDK API Reference*. 84 85The Link Bonding Library supports the creation of bonded devices at application startup time during EAL initialization using the 86--vdev option as well as programmatically via the C API rte_eth_bond_create function. 87 88Bonded devices support the dynamical addition and removal of slave devices using 89the rte_eth_bond_slave_add / rte_eth_bond_slave_remove APIs. 90 91After a slave device is added to a bonded device slave is stopped using 92rte_eth_dev_stop and the slave reconfigured using rte_eth_dev_configure the RX and TX queues are also reconfigured 93using rte_eth_tx_queue_setup / rte_eth_rx_queue_setup with the parameters use to configure the bonding device. 94 95Requirements / Limitations 96~~~~~~~~~~~~~~~~~~~~~~~~~~ 97 98The current implementation only supports physical devices of the same type, speed and duplex to be added as slaves. 99The bonded device inherits these values from the first active slave added to the bonded device 100and then all further slaves added to the bonded device must match these parameters. 101 102A bonding device must have a minimum of one slave before the bonding device itself can be started. 103 104Like all other PMD, all functions exported by a PMD are lock-free functions that are assumed 105not to be invoked in parallel on different logical cores to work on the same target object. 106 107It should also be noted that the PMD receive function should not be invoked directly on a slave devices after they have 108been to a bonded device since packets read directly from the slave device will no longer be available to the bonded device to read. 109 110Configuration 111~~~~~~~~~~~~~ 112 113Link bonding devices are created using the rte_eth_bond_create API 114which requires a unique device name, the bonding mode, 115and the socket Id to allocate the bonding device's resources on. 116The other configurable parameters for a bonded device are its slave devices, its primary slave, 117a user defined MAC address and transmission policy to use if the device is balance XOR mode. 118 119Slave Devices 120^^^^^^^^^^^^^ 121 122Bonding devices support up to a maximum of RTE_MAX_ETHPORTS slave devices of the same speed and duplex. 123Ethernet devices can be added as a slave to a maximum of one bonded device. 124Slave devices are reconfigured with the configuration of the bonded device on being added to a bonded device. 125 126The bonded also guarantees to return the MAC address of the slave device to its original value of removal of a slave from it. 127 128Primary Slave 129^^^^^^^^^^^^^ 130 131The primary slave is used to define the default port to use when a bonded device is in active backup mode. 132A different port will only be used if, and only if, the current primary port goes down. 133If the user does not specify a primary port it will default to being the first port added to the bonded device. 134 135MAC Address 136^^^^^^^^^^^ 137 138The bonded device can be configured with a user specified MAC address, 139this address will be inherited by the some/all slave devices depending on the operating mode. 140If the device is in active backup mode then only the primary device will have the user specified MAC, 141all other slaves will retain their original MAC address. 142In mode 0, 2, and 3 all slaves devices are configure with the bonded devices MAC address. 143 144If a user defined MAC address is not defined then the bonded device will default to using the primary slaves MAC address. 145 146Balance XOR Transmit Policies 147^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 148 149There are 3 supported transmission policies for bonded device running in Balance XOR mode. Layer 2, Layer 2+3, Layer 3+4. 150 151* **Layer 2:** Ethernet MAC address based balancing is the default transmission policy for Balance XOR bonding mode. 152 It uses a simple XOR calculation on the source MAC address and destination MAC address of the packet and 153 then calculate the modulus of this value to calculate the slave device to transmit the packet on. 154 155* **Layer 2 + 3:** Ethernet MAC address & IP Address based balancing uses a combination of source/destination MAC addresses and 156 the source/destination IP addresses of the data packet to decide which slave port the packet will be transmitted on. 157 158* **Layer 3 + 4:** IP Address & UDP Port based balancing uses a combination of source/destination IP Address and 159 the source/destination UDP ports of the packet of the data packet to decide which slave port the packet will be transmitted on. 160 161All these policies support 802.1Q VLAN Ethernet packets, as well as IPv4, IPv6 and UDP protocols for load balancing. 162 163Using Link Bonding Devices 164-------------------------- 165 166The librte_pmd_bond library support two modes of device creation, the libraries export full C API or 167using the EAL command line to statically configure link bonding devices at application startup. 168Using the EAL option it is possible to use link bonding functionality transparently without specific knowledge of the libraries API, 169this can be used, for example, to add bonding functionality, such as active backup, 170to an existing application which has no knowledge of the link bonding C API. 171 172Using the Poll Mode Driver from an Application 173~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 174 175Using the librte_pmd_bond libraries API it is possible to dynamicall create and manage link bonding device from within any application. 176Link bonding device are created using the rte_eth_bond_create API which requires a unqiue device name, 177the link bonding mode to initial the device in and finally the socket Id which to allocate the devices resources onto. 178After successful creation of a bonding device it must be configured using the generic Ethernet device configure API rte_eth_dev_configure 179and then the RX and TX queues which will be used must be setup using rte_eth_tx_queue_setup / rte_eth_rx_queue_setup. 180 181Slave devices can be dynamically added and removed from a link bonding device using the rte_eth_bond_slave_add / rte_eth_bond_slave_remove 182APIs but at least one slave device must be added to the link bonding device before it can be started using rte_eth_dev_start. 183 184The link status of a bonded device is dictated by that of its slaves, if all slave device link status are down or 185if all slaves are removed from the link bonding device then the link status of the bonding device will go down. 186 187It is also possible to configure / query the configuration of the control parameters of a bonded device using the provided APIs 188rte_eth_bond_mode_set/get, rte_eth_bond_primary_set/get, rte_eth_bond_mac_set/reset and rte_eth_bond_xmit_policy_set/get. 189 190Using Link Bonding Devices from the EAL Command Line 191~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 192 193Link bonding devices can be created at application startup time using the --vdev EAL command line option. 194The device name must start with the eth_bond prefix followed by numbers or letters. The name must be unique for each device. 195Each device can have multiple options arranged in a comma separated list. 196Multiple devices definitions can be arranged by calling the --vdev option multiple times. 197Device names and bonding options must be separated by commas as shown below: 198 199.. code-block:: console 200 201 $RTE_TARGET/app/testpmd -c f -n 4 --vdev 'eth_bond0,bond_opt0=..,bond opt1=..'--vdev 'eth_bond1,bond _opt0=..,bond_opt1=..' 202 203Link Bonding EAL Options 204^^^^^^^^^^^^^^^^^^^^^^^^ 205 206There are multiple ways of definitions that can be assessed and combined as long as the following two rules are respected: 207 208* A unique device name, in the format of eth_bondX is provided, 209 where X can be any combination of numbers and/or letters, 210 and the name is no greater than 32 characters long. 211 212* A least one slave device is provided with for each bonded device definition. 213 214* The operation mode of the bonded device being created is provided. 215 216The different options are: 217 218* mode: Integer value defining the bonding mode of the device. 219 Currently supports modes 0,1,2,3 (round-robin, active backup, balance, and broadcast). 220 221 mode=2 222 223* slave: Defines the PMD device which will be added as slave to the bonded device. 224 This option can be selected multiple time, for each device to be added as a slave. 225 Physical devices should be specified using their PCI address, in the format domain:bus:devid.function 226 227 slave=0000:0a:00.0,slave=0000:0a:00.1 228 229* primary: Optional parameter which defines the primary slave port, 230 is used in active backup mode to select the primary slave for data TX/RX if it is available. 231 The primary port also is used to select the MAC address to use when it is not defined by the user. 232 This defaults to the first slave added to the device if it is specified. 233 The primary device must be a slave of the bonded device. 234 235 primary=0000:0a:00.0 236 237* socket_id: Optional parameter used to select which socket on a NUMA device the bonded devices resources will be allocated on. 238 239 socket_id=0 240 241* mac: Optional parameter to select a MAC address for link bonding device, this overrides the value of the primary slave device. 242 243 mac=00:1e:67:1d:fd:1d 244 245* xmit_policy: Optional parameter which defines the transmission policy when the bonded device is in balance mode. 246 If not user specified this defaults to l2 (layer 2) forwarding, 247 the other transmission policies available are l23 (layer 2+3) and l34 (layer 3+4) 248 249 xmit_policy=l2 250 251Examples of Usage 252^^^^^^^^^^^^^^^^^ 253 254Create a bonded device in round robin mode with two slaves specified by their PCI address: 255 256.. code-block:: console 257 258 $RTE_TARGET/app/testpmd -c '0xf' -n 4 --vdev 'eth_bond0,mode=0, slave=0000:00a:00.01,slave=0000:004:00.00' -- --port-topology=chained 259 260Create a bonded device in round robin mode with two slaves specified by their PCI address and an overriding MAC address: 261 262.. code-block:: console 263 264 $RTE_TARGET/app/testpmd -c '0xf' -n 4 --vdev 'eth_bond0,mode=0, slave=0000:00a:00.01,slave=0000:004:00.00,mac=00:1e:67:1d:fd:1d' -- --port-topology=chained 265 266Create a bonded device in active backup mode with two slaves specified, and a primary slave specified by their PCI addresses: 267 268.. code-block:: console 269 270 $RTE_TARGET/app/testpmd -c '0xf' -n 4 --vdev 'eth_bond0,mode=1, slave=0000:00a:00.01,slave=0000:004:00.00,primary=0000:00a:00.01' -- --port-topology=chained 271 272Create a bonded device in balance mode with two slaves specified by their PCI addresses, and a transmission policy of layer 3 + 4 forwarding: 273 274.. code-block:: console 275 276 $RTE_TARGET/app/testpmd -c '0xf' -n 4 --vdev 'eth_bond0,mode=2, slave=0000:00a:00.01,slave=0000:004:00.00,xmit_policy=l34' -- --port-topology=chained 277 278.. |link_bonding| image:: img/link_bonding.png 279
