1 Using Open vSwitch with DPDK
2 ============================
4 Open vSwitch can use Intel(R) DPDK lib to operate entirely in
5 userspace. This file explains how to install and use Open vSwitch in
8 The DPDK support of Open vSwitch is considered experimental.
9 It has not been thoroughly tested.
11 This version of Open vSwitch should be built manually with `configure`
14 OVS needs a system with 1GB hugepages support.
16 Building and Installing:
17 ------------------------
21 1. Configure build & install DPDK:
25 export DPDK_DIR=/usr/src/dpdk-1.7.1
29 2. Update `config/common_linuxapp` so that DPDK generate single lib file.
30 (modification also required for IVSHMEM build)
32 `CONFIG_RTE_BUILD_COMBINE_LIBS=y`
34 Then run `make install` to build and isntall the library.
35 For default install without IVSHMEM:
37 `make install T=x86_64-native-linuxapp-gcc`
39 To include IVSHMEM (shared memory):
41 `make install T=x86_64-ivshmem-linuxapp-gcc`
43 For further details refer to http://dpdk.org/
45 2. Configure & build the Linux kernel:
47 Refer to intel-dpdk-getting-started-guide.pdf for understanding
48 DPDK kernel requirement.
50 3. Configure & build OVS:
54 `export DPDK_BUILD=$DPDK_DIR/x86_64-native-linuxapp-gcc/`
58 `export DPDK_BUILD=$DPDK_DIR/x86_64-ivshmem-linuxapp-gcc/`
61 cd $(OVS_DIR)/openvswitch
63 ./configure --with-dpdk=$DPDK_BUILD
67 To have better performance one can enable aggressive compiler optimizations and
68 use the special instructions(popcnt, crc32) that may not be available on all
69 machines. Instead of typing `make`, type:
71 `make CFLAGS='-O3 -march=native'`
73 Refer to [INSTALL.userspace](INSTALL.userspace.md) for general requirements of
74 building userspace OVS.
76 Using the DPDK with ovs-vswitchd:
77 ---------------------------------
80 Add the following options to the kernel bootline:
82 `default_hugepagesz=1GB hugepagesz=1G hugepages=1`
84 2. Setup DPDK devices:
85 1. insert uio.ko: `modprobe uio`
86 2. insert igb_uio.ko: `insmod $DPDK_BUILD/kmod/igb_uio.ko`
87 3. Bind network device to igb_uio: `$DPDK_DIR/tools/dpdk_nic_bind.py --bind=igb_uio eth1`
89 3. Mount the hugetable filsystem
91 `mount -t hugetlbfs -o pagesize=1G none /dev/hugepages`
93 Ref to http://www.dpdk.org/doc/quick-start for verifying DPDK setup.
95 4. Start ovsdb-server as discussed in [INSTALL](INSTALL.md) doc:
96 1. First time only db creation (or clearing):
99 mkdir -p /usr/local/etc/openvswitch
100 mkdir -p /usr/local/var/run/openvswitch
101 rm /usr/local/etc/openvswitch/conf.db
103 ./ovsdb/ovsdb-tool create /usr/local/etc/openvswitch/conf.db \
104 ./vswitchd/vswitch.ovsschema
107 2. start ovsdb-server
111 ./ovsdb/ovsdb-server --remote=punix:/usr/local/var/run/openvswitch/db.sock \
112 --remote=db:Open_vSwitch,Open_vSwitch,manager_options \
113 --private-key=db:Open_vSwitch,SSL,private_key \
114 --certificate=Open_vSwitch,SSL,certificate \
115 --bootstrap-ca-cert=db:Open_vSwitch,SSL,ca_cert --pidfile --detach
118 3. First time after db creation, initialize:
122 ./utilities/ovs-vsctl --no-wait init
127 DPDK configuration arguments can be passed to vswitchd via `--dpdk`
128 argument. This needs to be first argument passed to vswitchd process.
129 dpdk arg -c is ignored by ovs-dpdk, but it is a required parameter
130 for dpdk initialization.
132 export DB_SOCK=/usr/local/var/run/openvswitch/db.sock
133 ./vswitchd/ovs-vswitchd --dpdk -c 0x1 -n 4 -- unix:$DB_SOCK --pidfile --detach
135 If allocated more than one GB hugepage (as for IVSHMEM), set amount and use NUMA
138 ./vswitchd/ovs-vswitchd --dpdk -c 0x1 -n 4 --socket-mem 1024,0 \
139 -- unix:$DB_SOCK --pidfile --detach
141 6. Add bridge & ports
143 To use ovs-vswitchd with DPDK, create a bridge with datapath_type
144 "netdev" in the configuration database. For example:
146 `ovs-vsctl add-br br0 -- set bridge br0 datapath_type=netdev`
148 Now you can add dpdk devices. OVS expect DPDK device name start with dpdk
149 and end with portid. vswitchd should print (in the log file) the number of dpdk
152 ovs-vsctl add-port br0 dpdk0 -- set Interface dpdk0 type=dpdk
153 ovs-vsctl add-port br0 dpdk1 -- set Interface dpdk1 type=dpdk
155 Once first DPDK port is added to vswitchd, it creates a Polling thread and
156 polls dpdk device in continuous loop. Therefore CPU utilization
157 for that thread is always 100%.
161 Test flow script across NICs (assuming ovs in /usr/src/ovs):
166 # Move to command directory
167 cd /usr/src/ovs/utilities/
169 # Clear current flows
170 ./ovs-ofctl del-flows br0
172 # Add flows between port 1 (dpdk0) to port 2 (dpdk1)
173 ./ovs-ofctl add-flow br0 in_port=1,action=output:2
174 ./ovs-ofctl add-flow br0 in_port=2,action=output:1
177 8. Performance tuning
179 With pmd multi-threading support, OVS creates one pmd thread for each
180 numa node as default. The pmd thread handles the I/O of all DPDK
181 interfaces on the same numa node. The following two commands can be used
182 to configure the multi-threading behavior.
184 ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=<hex string>
186 The command above asks for a CPU mask for setting the affinity of pmd threads.
187 A set bit in the mask means a pmd thread is created and pinned to the
188 corresponding CPU core. For more information, please refer to
189 `man ovs-vswitchd.conf.db`
191 ovs-vsctl set Open_vSwitch . other_config:n-dpdk-rxqs=<integer>
193 The command above sets the number of rx queues of each DPDK interface. The
194 rx queues are assigned to pmd threads on the same numa node in round-robin
195 fashion. For more information, please refer to `man ovs-vswitchd.conf.db`
197 Ideally for maximum throughput, the pmd thread should not be scheduled out
198 which temporarily halts its execution. The following affinitization methods
201 Lets pick core 4,6,8,10 for pmd threads to run on. Also assume a dual 8 core
202 sandy bridge system with hyperthreading enabled where CPU1 has cores 0,...,7
203 and 16,...,23 & CPU2 cores 8,...,15 & 24,...,31. (A different cpu
204 configuration could have different core mask requirements).
206 To kernel bootline add core isolation list for cores and associated hype cores
207 (e.g. isolcpus=4,20,6,22,8,24,10,26,). Reboot system for isolation to take
208 effect, restart everything.
210 Configure pmd threads on core 4,6,8,10 using 'pmd-cpu-mask':
212 ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=00000550
214 You should be able to check that pmd threads are pinned to the correct cores
217 top -p `pidof ovs-vswitchd` -H -d1
219 Note, the pmd threads on a numa node are only created if there is at least
220 one DPDK interface from the numa node that has been added to OVS.
222 Note, core 0 is always reserved from non-pmd threads and should never be set
228 Following the steps above to create a bridge, you can now add dpdk rings
229 as a port to the vswitch. OVS will expect the DPDK ring device name to
230 start with dpdkr and end with a portid.
232 ovs-vsctl add-port br0 dpdkr0 -- set Interface dpdkr0 type=dpdkr
234 DPDK rings client test application
236 Included in the test directory is a sample DPDK application for testing
237 the rings. This is from the base dpdk directory and modified to work
238 with the ring naming used within ovs.
240 location tests/ovs_client
244 cd /usr/src/ovs/tests/
245 ovsclient -c 1 -n 4 --proc-type=secondary -- -n "port id you gave dpdkr"
247 In the case of the dpdkr example above the "port id you gave dpdkr" is 0.
249 It is essential to have --proc-type=secondary
251 The application simply receives an mbuf on the receive queue of the
252 ethernet ring and then places that same mbuf on the transmit ring of
253 the ethernet ring. It is a trivial loopback application.
255 DPDK rings in VM (IVSHMEM shared memory communications)
256 -------------------------------------------------------
258 In addition to executing the client in the host, you can execute it within
259 a guest VM. To do so you will need a patched qemu. You can download the
260 patch and getting started guide at :
262 https://01.org/packet-processing/downloads
264 A general rule of thumb for better performance is that the client
265 application should not be assigned the same dpdk core mask "-c" as
271 - This Support is for Physical NIC. I have tested with Intel NIC only.
272 - Work with 1500 MTU, needs few changes in DPDK lib to fix this issue.
273 - Currently DPDK port does not make use any offload functionality.
276 - The shared memory is currently restricted to the use of a 1GB
278 - All huge pages are shared amongst the host, clients, virtual
284 Please report problems to bugs@openvswitch.org.