| Patrick Williams | d849ec7 | 2016-08-17 14:59:38 -0500 | [diff] [blame] | 1 | Simple setup for connecting openvswitch to qemu/kvm |
| 2 | =================================================== |
| 3 | This example brings up openvswitch using a private network. |
| 4 | |
| 5 | Preliminary notes |
| 6 | ================= |
| 7 | 1. Make sure to build kernel support for openvswitch as a module. The |
| 8 | openvswitch init scripts expect to load a module and upon success |
| 9 | continue to setup the switch. If openvswitch is compiled |
| 10 | statically, the init scripts not load the ovs-vswitchd daemon |
| 11 | and none of the configured bridges will show up in the interfaces |
| 12 | table (ifconfig). You can get around this limiation by running the |
| 13 | following by hand: |
| 14 | # ovs-vswitchd --pidfile --detach |
| 15 | |
| 16 | 2. Verify that ovs-vswitchd is running before proceeding: |
| 17 | # /etc/init.d/openvswitch-switch status |
| 18 | ovsdb-server is running with pid 1867 |
| 19 | ovs-vswitchd is running with pid 1877 |
| 20 | |
| 21 | 3. A kernel and rootfs is required for qemu bring up. |
| 22 | |
| 23 | Qemu Setup |
| 24 | ========== |
| 25 | The host requires a /etc/qemu-ifup script to setup the bridging and tap |
| 26 | devices. Qemu will invoke this qemu-ifup script at startup. Here is |
| 27 | an example script: |
| 28 | $ cat /etc/qemu-fup |
| 29 | #!/bin/sh |
| 30 | # the tap is dynamically assigned and passed into this script |
| 31 | # as a parameter |
| 32 | TAP=$1 |
| 33 | |
| 34 | # Note: if booting over NFS, once the $ETH0 device is added to the bridge, |
| 35 | # your host will be unusable. In that case, setup networking |
| 36 | # init scripts appropriately and change the following to work |
| 37 | # with it. |
| 38 | ETH0="eth1" |
| 39 | NETMASK=255.255.255.0 |
| 40 | IP=192.168.1.1 |
| 41 | GATEWAY= |
| 42 | SWITCH=ovsbr0 |
| 43 | if [ -n "$TAP" ];then |
| 44 | ifconfig $TAP up |
| 45 | ifconfig $SWITCH down &>/dev/null |
| 46 | ovs-vsctl del-br $SWITCH |
| 47 | ovs-vsctl add-br $SWITCH |
| 48 | ifconfig $ETH0 0.0.0.0 |
| 49 | ifconfig $SWITCH $IP up netmask $NETMASK |
| 50 | #-- external access not required for this test. |
| 51 | #route add default gw $GATEWAY |
| 52 | ovs-vsctl add-port $SWITCH $ETH0 |
| 53 | ovs-vsctl add-port $SWITCH $TAP |
| 54 | exit 0 |
| 55 | else |
| 56 | echo "$0: No tap device" |
| 57 | exit 1 |
| 58 | fi |
| 59 | |
| 60 | Start Qemu |
| 61 | ========== |
| 62 | This example will bring up qemu with a tap network interface. |
| 63 | Note: this command must be run as root due to the networking setup. |
| 64 | |
| 65 | $ qemu-system-x86_64 -nographic -k en-us -m 1024 \ |
| 66 | -net nic,macaddr=1a:46:0b:ca:bc:7a,model=virtio \ |
| 67 | -net tap -enable-kvm\ |
| 68 | -kernel /opt/dpdk-guest-kernel \ |
| 69 | -append 'root=/dev/vda ro console=ttyS0' \ |
| 70 | -drive file=/opt/intel-xeon-core-ovp-kvm-preempt-rt-dist.ext3,cache=none,if=virtio |
| 71 | |
| 72 | Once the guest OS is up and running, configure the quest network interface: |
| 73 | $ ifconfig eth0 192.168.1.10 |
| 74 | |
| 75 | Ping the bridge: |
| 76 | $ ping 192.168.1.1 |
| 77 | |
| 78 | From the host, view the bridged network: |
| 79 | $ ovs-vsctl show |
| 80 | c1212b96-ef49-4a8e-b598-09b05b854dd0 |
| 81 | Bridge "ovsbr0" |
| 82 | Port "tap0" |
| 83 | Interface "tap0" |
| 84 | Port "eth1" |
| 85 | Interface "eth1" |
| 86 | Port "ovsbr0" |
| 87 | Interface "ovsbr0" |
| 88 | type: internal |
| 89 | |
| 90 | At this point, openvswitch is up and running. If you want external |
| 91 | network access, you need to set a GATEWAY in the qemu-ifup script and |
| 92 | make sure the external device is part of the bridge. |
| 93 | |
| 94 | Note: |
| 95 | Proper setup will require a /etc/qemu-ifdown script to tear down the |
| 96 | bridge and interfaces. (not provided here). |