1 Open vSwitch <http://openvswitch.org>
3 Frequently Asked Questions
4 ==========================
9 Q: What is Open vSwitch?
11 A: Open vSwitch is a production quality open source software switch
12 designed to be used as a vswitch in virtualized server
13 environments. A vswitch forwards traffic between different VMs on
14 the same physical host and also forwards traffic between VMs and
15 the physical network. Open vSwitch supports standard management
16 interfaces (e.g. sFlow, NetFlow, IPFIX, RSPAN, CLI), and is open to
17 programmatic extension and control using OpenFlow and the OVSDB
20 Open vSwitch as designed to be compatible with modern switching
21 chipsets. This means that it can be ported to existing high-fanout
22 switches allowing the same flexible control of the physical
23 infrastructure as the virtual infrastructure. It also means that
24 Open vSwitch will be able to take advantage of on-NIC switching
25 chipsets as their functionality matures.
27 Q: What virtualization platforms can use Open vSwitch?
29 A: Open vSwitch can currently run on any Linux-based virtualization
30 platform (kernel 2.6.32 and newer), including: KVM, VirtualBox, Xen,
31 Xen Cloud Platform, XenServer. As of Linux 3.3 it is part of the
32 mainline kernel. The bulk of the code is written in platform-
33 independent C and is easily ported to other environments. We welcome
34 inquires about integrating Open vSwitch with other virtualization
37 Q: How can I try Open vSwitch?
39 A: The Open vSwitch source code can be built on a Linux system. You can
40 build and experiment with Open vSwitch on any Linux machine.
41 Packages for various Linux distributions are available on many
42 platforms, including: Debian, Ubuntu, Fedora.
44 You may also download and run a virtualization platform that already
45 has Open vSwitch integrated. For example, download a recent ISO for
46 XenServer or Xen Cloud Platform. Be aware that the version
47 integrated with a particular platform may not be the most recent Open
50 Q: Does Open vSwitch only work on Linux?
52 A: No, Open vSwitch has been ported to a number of different operating
53 systems and hardware platforms. Most of the development work occurs
54 on Linux, but the code should be portable to any POSIX system. We've
55 seen Open vSwitch ported to a number of different platforms,
56 including FreeBSD, Windows, and even non-POSIX embedded systems.
58 By definition, the Open vSwitch Linux kernel module only works on
59 Linux and will provide the highest performance. However, a userspace
60 datapath is available that should be very portable.
62 Q: What's involved with porting Open vSwitch to a new platform or
65 A: The PORTING document describes how one would go about porting Open
66 vSwitch to a new operating system or hardware platform.
68 Q: Why would I use Open vSwitch instead of the Linux bridge?
70 A: Open vSwitch is specially designed to make it easier to manage VM
71 network configuration and monitor state spread across many physical
72 hosts in dynamic virtualized environments. Please see WHY-OVS for a
73 more detailed description of how Open vSwitch relates to the Linux
76 Q: How is Open vSwitch related to distributed virtual switches like the
77 VMware vNetwork distributed switch or the Cisco Nexus 1000V?
79 A: Distributed vswitch applications (e.g., VMware vNetwork distributed
80 switch, Cisco Nexus 1000V) provide a centralized way to configure and
81 monitor the network state of VMs that are spread across many physical
82 hosts. Open vSwitch is not a distributed vswitch itself, rather it
83 runs on each physical host and supports remote management in a way
84 that makes it easier for developers of virtualization/cloud
85 management platforms to offer distributed vswitch capabilities.
87 To aid in distribution, Open vSwitch provides two open protocols that
88 are specially designed for remote management in virtualized network
89 environments: OpenFlow, which exposes flow-based forwarding state,
90 and the OVSDB management protocol, which exposes switch port state.
91 In addition to the switch implementation itself, Open vSwitch
92 includes tools (ovs-controller, ovs-ofctl, ovs-vsctl) that developers
93 can script and extend to provide distributed vswitch capabilities
94 that are closely integrated with their virtualization management
97 Q: Why doesn't Open vSwitch support distribution?
99 A: Open vSwitch is intended to be a useful component for building
100 flexible network infrastructure. There are many different approaches
101 to distribution which balance trade-offs between simplicity,
102 scalability, hardware compatibility, convergence times, logical
103 forwarding model, etc. The goal of Open vSwitch is to be able to
104 support all as a primitive building block rather than choose a
105 particular point in the distributed design space.
107 Q: How can I contribute to the Open vSwitch Community?
109 A: You can start by joining the mailing lists and helping to answer
110 questions. You can also suggest improvements to documentation. If
111 you have a feature or bug you would like to work on, send a mail to
112 one of the mailing lists:
114 http://openvswitch.org/mlists/
120 Q: What does it mean for an Open vSwitch release to be LTS (long-term
123 A: All official releases have been through a comprehensive testing
124 process and are suitable for production use. Planned releases will
125 occur several times a year. If a significant bug is identified in an
126 LTS release, we will provide an updated release that includes the
127 fix. Releases that are not LTS may not be fixed and may just be
128 supplanted by the next major release. The current LTS release is
131 Q: What Linux kernel versions does each Open vSwitch release work with?
133 A: The following table lists the Linux kernel versions against which the
134 given versions of the Open vSwitch kernel module will successfully
135 build. The Linux kernel versions are upstream kernel versions, so
136 Linux kernels modified from the upstream sources may not build in
137 some cases even if they are based on a supported version. This is
138 most notably true of Red Hat Enterprise Linux (RHEL) kernels, which
139 are extensively modified from upstream.
141 Open vSwitch Linux kernel
142 ------------ -------------
154 Open vSwitch userspace should also work with the Linux kernel module
155 built into Linux 3.3 and later.
157 Open vSwitch userspace is not sensitive to the Linux kernel version.
158 It should build against almost any kernel, certainly against 2.6.32
161 Q: What Linux kernel versions does IPFIX flow monitoring work with?
163 A: IPFIX flow monitoring requires the Linux kernel module from Open
164 vSwitch version 1.10.90 or later.
166 Q: Should userspace or kernel be upgraded first to minimize downtime?
168 In general, the Open vSwitch userspace should be used with the
169 kernel version included in the same release or with the version
170 from upstream Linux. However, when upgrading between two releases
171 of Open vSwitch it is best to migrate userspace first to reduce
172 the possibility of incompatibilities.
174 Q: What features are not available in the Open vSwitch kernel datapath
175 that ships as part of the upstream Linux kernel?
177 A: The kernel module in upstream Linux 3.3 and later does not include
178 tunnel virtual ports, that is, interfaces with type "gre",
179 "ipsec_gre", "gre64", "ipsec_gre64", "vxlan", or "lisp". It is
180 possible to create tunnels in Linux and attach them to Open vSwitch
181 as system devices. However, they cannot be dynamically created
182 through the OVSDB protocol or set the tunnel ids as a flow action.
184 Work is in progress in adding tunnel virtual ports to the upstream
185 Linux version of the Open vSwitch kernel module. For now, if you
186 need these features, use the kernel module from the Open vSwitch
187 distribution instead of the upstream Linux kernel module.
189 The upstream kernel module does not include patch ports, but this
190 only matters for Open vSwitch 1.9 and earlier, because Open vSwitch
191 1.10 and later implement patch ports without using this kernel
194 Q: What features are not available when using the userspace datapath?
196 A: Tunnel virtual ports are not supported, as described in the
197 previous answer. It is also not possible to use queue-related
198 actions. On Linux kernels before 2.6.39, maximum-sized VLAN packets
199 may not be transmitted.
201 Q: What happened to the bridge compatibility feature?
203 A: Bridge compatibility was a feature of Open vSwitch 1.9 and earlier.
204 When it was enabled, Open vSwitch imitated the interface of the
205 Linux kernel "bridge" module. This allowed users to drop Open
206 vSwitch into environments designed to use the Linux kernel bridge
207 module without adapting the environment to use Open vSwitch.
209 Open vSwitch 1.10 and later do not support bridge compatibility.
210 The feature was dropped because version 1.10 adopted a new internal
211 architecture that made bridge compatibility difficult to maintain.
212 Now that many environments use OVS directly, it would be rarely
215 To use bridge compatibility, install OVS 1.9 or earlier, including
216 the accompanying kernel modules (both the main and bridge
217 compatibility modules), following the instructions that come with
218 the release. Be sure to start the ovs-brcompatd daemon.
224 Q: I thought Open vSwitch was a virtual Ethernet switch, but the
225 documentation keeps talking about bridges. What's a bridge?
227 A: In networking, the terms "bridge" and "switch" are synonyms. Open
228 vSwitch implements an Ethernet switch, which means that it is also
233 A: See the "VLAN" section below.
239 Q: How do I configure a port as an access port?
241 A: Add "tag=VLAN" to your "ovs-vsctl add-port" command. For example,
242 the following commands configure br0 with eth0 as a trunk port (the
243 default) and tap0 as an access port for VLAN 9:
246 ovs-vsctl add-port br0 eth0
247 ovs-vsctl add-port br0 tap0 tag=9
249 If you want to configure an already added port as an access port,
250 use "ovs-vsctl set", e.g.:
252 ovs-vsctl set port tap0 tag=9
254 Q: How do I configure a port as a SPAN port, that is, enable mirroring
255 of all traffic to that port?
257 A: The following commands configure br0 with eth0 and tap0 as trunk
258 ports. All traffic coming in or going out on eth0 or tap0 is also
259 mirrored to tap1; any traffic arriving on tap1 is dropped:
262 ovs-vsctl add-port br0 eth0
263 ovs-vsctl add-port br0 tap0
264 ovs-vsctl add-port br0 tap1 \
265 -- --id=@p get port tap1 \
266 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
267 -- set bridge br0 mirrors=@m
269 To later disable mirroring, run:
271 ovs-vsctl clear bridge br0 mirrors
273 Q: Does Open vSwitch support configuring a port in promiscuous mode?
275 A: Yes. How you configure it depends on what you mean by "promiscuous
278 - Conventionally, "promiscuous mode" is a feature of a network
279 interface card. Ordinarily, a NIC passes to the CPU only the
280 packets actually destined to its host machine. It discards
281 the rest to avoid wasting memory and CPU cycles. When
282 promiscuous mode is enabled, however, it passes every packet
283 to the CPU. On an old-style shared-media or hub-based
284 network, this allows the host to spy on all packets on the
285 network. But in the switched networks that are almost
286 everywhere these days, promiscuous mode doesn't have much
287 effect, because few packets not destined to a host are
288 delivered to the host's NIC.
290 This form of promiscuous mode is configured in the guest OS of
291 the VMs on your bridge, e.g. with "ifconfig".
293 - The VMware vSwitch uses a different definition of "promiscuous
294 mode". When you configure promiscuous mode on a VMware vNIC,
295 the vSwitch sends a copy of every packet received by the
296 vSwitch to that vNIC. That has a much bigger effect than just
297 enabling promiscuous mode in a guest OS. Rather than getting
298 a few stray packets for which the switch does not yet know the
299 correct destination, the vNIC gets every packet. The effect
300 is similar to replacing the vSwitch by a virtual hub.
302 This "promiscuous mode" is what switches normally call "port
303 mirroring" or "SPAN". For information on how to configure
304 SPAN, see "How do I configure a port as a SPAN port, that is,
305 enable mirroring of all traffic to that port?"
307 Q: How do I configure a VLAN as an RSPAN VLAN, that is, enable
308 mirroring of all traffic to that VLAN?
310 A: The following commands configure br0 with eth0 as a trunk port and
311 tap0 as an access port for VLAN 10. All traffic coming in or going
312 out on tap0, as well as traffic coming in or going out on eth0 in
313 VLAN 10, is also mirrored to VLAN 15 on eth0. The original tag for
314 VLAN 10, in cases where one is present, is dropped as part of
318 ovs-vsctl add-port br0 eth0
319 ovs-vsctl add-port br0 tap0 tag=10
321 -- --id=@m create mirror name=m0 select-all=true select-vlan=10 \
323 -- set bridge br0 mirrors=@m
325 To later disable mirroring, run:
327 ovs-vsctl clear bridge br0 mirrors
329 Mirroring to a VLAN can disrupt a network that contains unmanaged
330 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
331 GRE tunnel has fewer caveats than mirroring to a VLAN and should
332 generally be preferred.
334 Q: Can I mirror more than one input VLAN to an RSPAN VLAN?
336 A: Yes, but mirroring to a VLAN strips the original VLAN tag in favor
337 of the specified output-vlan. This loss of information may make
338 the mirrored traffic too hard to interpret.
340 To mirror multiple VLANs, use the commands above, but specify a
341 comma-separated list of VLANs as the value for select-vlan. To
342 mirror every VLAN, use the commands above, but omit select-vlan and
345 When a packet arrives on a VLAN that is used as a mirror output
346 VLAN, the mirror is disregarded. Instead, in standalone mode, OVS
347 floods the packet across all the ports for which the mirror output
348 VLAN is configured. (If an OpenFlow controller is in use, then it
349 can override this behavior through the flow table.) If OVS is used
350 as an intermediate switch, rather than an edge switch, this ensures
351 that the RSPAN traffic is distributed through the network.
353 Mirroring to a VLAN can disrupt a network that contains unmanaged
354 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
355 GRE tunnel has fewer caveats than mirroring to a VLAN and should
356 generally be preferred.
358 Q: How do I configure mirroring of all traffic to a GRE tunnel?
360 A: The following commands configure br0 with eth0 and tap0 as trunk
361 ports. All traffic coming in or going out on eth0 or tap0 is also
362 mirrored to gre0, a GRE tunnel to the remote host 192.168.1.10; any
363 traffic arriving on gre0 is dropped:
366 ovs-vsctl add-port br0 eth0
367 ovs-vsctl add-port br0 tap0
368 ovs-vsctl add-port br0 gre0 \
369 -- set interface gre0 type=gre options:remote_ip=192.168.1.10 \
370 -- --id=@p get port gre0 \
371 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
372 -- set bridge br0 mirrors=@m
374 To later disable mirroring and destroy the GRE tunnel:
376 ovs-vsctl clear bridge br0 mirrors
377 ovs-vcstl del-port br0 gre0
379 Q: Does Open vSwitch support ERSPAN?
381 A: No. ERSPAN is an undocumented proprietary protocol. As an
382 alternative, Open vSwitch supports mirroring to a GRE tunnel (see
385 Q: How do I connect two bridges?
387 A: First, why do you want to do this? Two connected bridges are not
388 much different from a single bridge, so you might as well just have
389 a single bridge with all your ports on it.
391 If you still want to connect two bridges, you can use a pair of
392 patch ports. The following example creates bridges br0 and br1,
393 adds eth0 and tap0 to br0, adds tap1 to br1, and then connects br0
394 and br1 with a pair of patch ports.
397 ovs-vsctl add-port br0 eth0
398 ovs-vsctl add-port br0 tap0
400 ovs-vsctl add-port br1 tap1
402 -- add-port br0 patch0 \
403 -- set interface patch0 type=patch options:peer=patch1 \
404 -- add-port br1 patch1 \
405 -- set interface patch1 type=patch options:peer=patch0
407 Bridges connected with patch ports are much like a single bridge.
408 For instance, if the example above also added eth1 to br1, and both
409 eth0 and eth1 happened to be connected to the same next-hop switch,
410 then you could loop your network just as you would if you added
411 eth0 and eth1 to the same bridge (see the "Configuration Problems"
412 section below for more information).
414 If you are using Open vSwitch 1.9 or an earlier version, then you
415 need to be using the kernel module bundled with Open vSwitch rather
416 than the one that is integrated into Linux 3.3 and later, because
417 Open vSwitch 1.9 and earlier versions need kernel support for patch
418 ports. This also means that in Open vSwitch 1.9 and earlier, patch
419 ports will not work with the userspace datapath, only with the
422 Q: Why are there so many different ways to dump flows?
424 A: Open vSwitch uses different kinds of flows for different purposes:
426 - OpenFlow flows are the most important kind of flow. OpenFlow
427 controllers use these flows to define a switch's policy.
428 OpenFlow flows support wildcards, priorities, and multiple
431 When in-band control is in use, Open vSwitch sets up a few
432 "hidden" flows, with priority higher than a controller or the
433 user can configure, that are not visible via OpenFlow. (See
434 the "Controller" section of the FAQ for more information
437 - The Open vSwitch software switch implementation uses a second
438 kind of flow internally. These flows, called "exact-match"
439 or "datapath" or "kernel" flows, do not support wildcards or
440 priorities and comprise only a single table, which makes them
441 suitable for caching. OpenFlow flows and exact-match flows
442 also support different actions and number ports differently.
444 Exact-match flows are an implementation detail that is
445 subject to change in future versions of Open vSwitch. Even
446 with the current version of Open vSwitch, hardware switch
447 implementations do not necessarily use exact-match flows.
449 Each of the commands for dumping flows has a different purpose:
451 - "ovs-ofctl dump-flows <br>" dumps OpenFlow flows, excluding
452 hidden flows. This is the most commonly useful form of flow
453 dump. (Unlike the other commands, this should work with any
454 OpenFlow switch, not just Open vSwitch.)
456 - "ovs-appctl bridge/dump-flows <br>" dumps OpenFlow flows,
457 including hidden flows. This is occasionally useful for
458 troubleshooting suspected issues with in-band control.
460 - "ovs-dpctl dump-flows [dp]" dumps the exact-match flow table
461 entries for a Linux kernel-based datapath. In Open vSwitch
462 1.10 and later, ovs-vswitchd merges multiple switches into a
463 single datapath, so it will show all the flows on all your
464 kernel-based switches. This command can occasionally be
465 useful for debugging.
467 - "ovs-appctl dpif/dump-flows <br>", new in Open vSwitch 1.10,
468 dumps exact-match flows for only the specified bridge,
469 regardless of the type.
472 Configuration Problems
473 ----------------------
475 Q: I created a bridge and added my Ethernet port to it, using commands
479 ovs-vsctl add-port br0 eth0
481 and as soon as I ran the "add-port" command I lost all connectivity
484 A: A physical Ethernet device that is part of an Open vSwitch bridge
485 should not have an IP address. If one does, then that IP address
486 will not be fully functional.
488 You can restore functionality by moving the IP address to an Open
489 vSwitch "internal" device, such as the network device named after
490 the bridge itself. For example, assuming that eth0's IP address is
491 192.168.128.5, you could run the commands below to fix up the
494 ifconfig eth0 0.0.0.0
495 ifconfig br0 192.168.128.5
497 (If your only connection to the machine running OVS is through the
498 IP address in question, then you would want to run all of these
499 commands on a single command line, or put them into a script.) If
500 there were any additional routes assigned to eth0, then you would
501 also want to use commands to adjust these routes to go through br0.
503 If you use DHCP to obtain an IP address, then you should kill the
504 DHCP client that was listening on the physical Ethernet interface
505 (e.g. eth0) and start one listening on the internal interface
506 (e.g. br0). You might still need to manually clear the IP address
507 from the physical interface (e.g. with "ifconfig eth0 0.0.0.0").
509 There is no compelling reason why Open vSwitch must work this way.
510 However, this is the way that the Linux kernel bridge module has
511 always worked, so it's a model that those accustomed to Linux
512 bridging are already used to. Also, the model that most people
513 expect is not implementable without kernel changes on all the
514 versions of Linux that Open vSwitch supports.
516 By the way, this issue is not specific to physical Ethernet
517 devices. It applies to all network devices except Open vswitch
520 Q: I created a bridge and added a couple of Ethernet ports to it,
521 using commands like these:
524 ovs-vsctl add-port br0 eth0
525 ovs-vsctl add-port br0 eth1
527 and now my network seems to have melted: connectivity is unreliable
528 (even connectivity that doesn't go through Open vSwitch), all the
529 LEDs on my physical switches are blinking, wireshark shows
530 duplicated packets, and CPU usage is very high.
532 A: More than likely, you've looped your network. Probably, eth0 and
533 eth1 are connected to the same physical Ethernet switch. This
534 yields a scenario where OVS receives a broadcast packet on eth0 and
535 sends it out on eth1, then the physical switch connected to eth1
536 sends the packet back on eth0, and so on forever. More complicated
537 scenarios, involving a loop through multiple switches, are possible
540 The solution depends on what you are trying to do:
542 - If you added eth0 and eth1 to get higher bandwidth or higher
543 reliability between OVS and your physical Ethernet switch,
544 use a bond. The following commands create br0 and then add
545 eth0 and eth1 as a bond:
548 ovs-vsctl add-bond br0 bond0 eth0 eth1
550 Bonds have tons of configuration options. Please read the
551 documentation on the Port table in ovs-vswitchd.conf.db(5)
554 - Perhaps you don't actually need eth0 and eth1 to be on the
555 same bridge. For example, if you simply want to be able to
556 connect each of them to virtual machines, then you can put
557 each of them on a bridge of its own:
560 ovs-vsctl add-port br0 eth0
563 ovs-vsctl add-port br1 eth1
565 and then connect VMs to br0 and br1. (A potential
566 disadvantage is that traffic cannot directly pass between br0
567 and br1. Instead, it will go out eth0 and come back in eth1,
570 - If you have a redundant or complex network topology and you
571 want to prevent loops, turn on spanning tree protocol (STP).
572 The following commands create br0, enable STP, and add eth0
573 and eth1 to the bridge. The order is important because you
574 don't want have to have a loop in your network even
578 ovs-vsctl set bridge br0 stp_enable=true
579 ovs-vsctl add-port br0 eth0
580 ovs-vsctl add-port br0 eth1
582 The Open vSwitch implementation of STP is not well tested.
583 Please report any bugs you observe, but if you'd rather avoid
584 acting as a beta tester then another option might be your
587 Q: I can't seem to use Open vSwitch in a wireless network.
589 A: Wireless base stations generally only allow packets with the source
590 MAC address of NIC that completed the initial handshake.
591 Therefore, without MAC rewriting, only a single device can
592 communicate over a single wireless link.
594 This isn't specific to Open vSwitch, it's enforced by the access
595 point, so the same problems will show up with the Linux bridge or
596 any other way to do bridging.
598 Q: I can't seem to add my PPP interface to an Open vSwitch bridge.
600 A: PPP most commonly carries IP packets, but Open vSwitch works only
601 with Ethernet frames. The correct way to interface PPP to an
602 Ethernet network is usually to use routing instead of switching.
604 Q: Is there any documentation on the database tables and fields?
606 A: Yes. ovs-vswitchd.conf.db(5) is a comprehensive reference.
608 Q: When I run ovs-dpctl I no longer see the bridges I created. Instead,
609 I only see a datapath called "ovs-system". How can I see datapath
610 information about a particular bridge?
612 A: In version 1.9.0, OVS switched to using a single datapath that is
613 shared by all bridges of that type. The "ovs-appctl dpif/*"
614 commands provide similar functionality that is scoped by the bridge.
617 Quality of Service (QoS)
618 ------------------------
620 Q: How do I configure Quality of Service (QoS)?
622 A: Suppose that you want to set up bridge br0 connected to physical
623 Ethernet port eth0 (a 1 Gbps device) and virtual machine interfaces
624 vif1.0 and vif2.0, and that you want to limit traffic from vif1.0
625 to eth0 to 10 Mbps and from vif2.0 to eth0 to 20 Mbps. Then, you
626 could configure the bridge this way:
630 add-port br0 eth0 -- \
631 add-port br0 vif1.0 -- set interface vif1.0 ofport_request=5 -- \
632 add-port br0 vif2.0 -- set interface vif2.0 ofport_request=6 -- \
633 set port eth0 qos=@newqos -- \
634 --id=@newqos create qos type=linux-htb \
635 other-config:max-rate=1000000000 \
636 queues:123=@vif10queue \
637 queues:234=@vif20queue -- \
638 --id=@vif10queue create queue other-config:max-rate=10000000 -- \
639 --id=@vif20queue create queue other-config:max-rate=20000000
641 At this point, bridge br0 is configured with the ports and eth0 is
642 configured with the queues that you need for QoS, but nothing is
643 actually directing packets from vif1.0 or vif2.0 to the queues that
644 we have set up for them. That means that all of the packets to
645 eth0 are going to the "default queue", which is not what we want.
647 We use OpenFlow to direct packets from vif1.0 and vif2.0 to the
648 queues reserved for them:
650 ovs-ofctl add-flow br0 in_port=5,actions=set_queue:123,normal
651 ovs-ofctl add-flow br0 in_port=6,actions=set_queue:234,normal
653 Each of the above flows matches on the input port, sets up the
654 appropriate queue (123 for vif1.0, 234 for vif2.0), and then
655 executes the "normal" action, which performs the same switching
656 that Open vSwitch would have done without any OpenFlow flows being
657 present. (We know that vif1.0 and vif2.0 have OpenFlow port
658 numbers 5 and 6, respectively, because we set their ofport_request
659 columns above. If we had not done that, then we would have needed
660 to find out their port numbers before setting up these flows.)
662 Now traffic going from vif1.0 or vif2.0 to eth0 should be
665 By the way, if you delete the bridge created by the above commands,
670 then that will leave one unreferenced QoS record and two
671 unreferenced Queue records in the Open vSwich database. One way to
672 clear them out, assuming you don't have other QoS or Queue records
673 that you want to keep, is:
675 ovs-vsctl -- --all destroy QoS -- --all destroy Queue
677 If you do want to keep some QoS or Queue records, or the Open
678 vSwitch you are using is older than version 1.8 (which added the
679 --all option), then you will have to destroy QoS and Queue records
682 Q: I configured Quality of Service (QoS) in my OpenFlow network by
683 adding records to the QoS and Queue table, but the results aren't
686 A: Did you install OpenFlow flows that use your queues? This is the
687 primary way to tell Open vSwitch which queues you want to use. If
688 you don't do this, then the default queue will be used, which will
689 probably not have the effect you want.
691 Refer to the previous question for an example.
693 Q: I configured QoS, correctly, but my measurements show that it isn't
694 working as well as I expect.
696 A: With the Linux kernel, the Open vSwitch implementation of QoS has
699 - Open vSwitch configures a subset of Linux kernel QoS
700 features, according to what is in OVSDB. It is possible that
701 this code has bugs. If you believe that this is so, then you
702 can configure the Linux traffic control (QoS) stack directly
703 with the "tc" program. If you get better results that way,
704 you can send a detailed bug report to bugs@openvswitch.org.
706 It is certain that Open vSwitch cannot configure every Linux
707 kernel QoS feature. If you need some feature that OVS cannot
708 configure, then you can also use "tc" directly (or add that
711 - The Open vSwitch implementation of OpenFlow allows flows to
712 be directed to particular queues. This is pretty simple and
713 unlikely to have serious bugs at this point.
715 However, most problems with QoS on Linux are not bugs in Open
716 vSwitch at all. They tend to be either configuration errors
717 (please see the earlier questions in this section) or issues with
718 the traffic control (QoS) stack in Linux. The Open vSwitch
719 developers are not experts on Linux traffic control. We suggest
720 that, if you believe you are encountering a problem with Linux
721 traffic control, that you consult the tc manpages (e.g. tc(8),
722 tc-htb(8), tc-hfsc(8)), web resources (e.g. http://lartc.org/), or
723 mailing lists (e.g. http://vger.kernel.org/vger-lists.html#netdev).
731 A: At the simplest level, a VLAN (short for "virtual LAN") is a way to
732 partition a single switch into multiple switches. Suppose, for
733 example, that you have two groups of machines, group A and group B.
734 You want the machines in group A to be able to talk to each other,
735 and you want the machine in group B to be able to talk to each
736 other, but you don't want the machines in group A to be able to
737 talk to the machines in group B. You can do this with two
738 switches, by plugging the machines in group A into one switch and
739 the machines in group B into the other switch.
741 If you only have one switch, then you can use VLANs to do the same
742 thing, by configuring the ports for machines in group A as VLAN
743 "access ports" for one VLAN and the ports for group B as "access
744 ports" for a different VLAN. The switch will only forward packets
745 between ports that are assigned to the same VLAN, so this
746 effectively subdivides your single switch into two independent
747 switches, one for each group of machines.
749 So far we haven't said anything about VLAN headers. With access
750 ports, like we've described so far, no VLAN header is present in
751 the Ethernet frame. This means that the machines (or switches)
752 connected to access ports need not be aware that VLANs are
753 involved, just like in the case where we use two different physical
756 Now suppose that you have a whole bunch of switches in your
757 network, instead of just one, and that some machines in group A are
758 connected directly to both switches 1 and 2. To allow these
759 machines to talk to each other, you could add an access port for
760 group A's VLAN to switch 1 and another to switch 2, and then
761 connect an Ethernet cable between those ports. That works fine,
762 but it doesn't scale well as the number of switches and the number
763 of VLANs increases, because you use up a lot of valuable switch
764 ports just connecting together your VLANs.
766 This is where VLAN headers come in. Instead of using one cable and
767 two ports per VLAN to connect a pair of switches, we configure a
768 port on each switch as a VLAN "trunk port". Packets sent and
769 received on a trunk port carry a VLAN header that says what VLAN
770 the packet belongs to, so that only two ports total are required to
771 connect the switches, regardless of the number of VLANs in use.
772 Normally, only switches (either physical or virtual) are connected
773 to a trunk port, not individual hosts, because individual hosts
774 don't expect to see a VLAN header in the traffic that they receive.
776 None of the above discussion says anything about particular VLAN
777 numbers. This is because VLAN numbers are completely arbitrary.
778 One must only ensure that a given VLAN is numbered consistently
779 throughout a network and that different VLANs are given different
780 numbers. (That said, VLAN 0 is usually synonymous with a packet
781 that has no VLAN header, and VLAN 4095 is reserved.)
785 A: Many drivers in Linux kernels before version 3.3 had VLAN-related
786 bugs. If you are having problems with VLANs that you suspect to be
787 driver related, then you have several options:
789 - Upgrade to Linux 3.3 or later.
791 - Build and install a fixed version of the particular driver
792 that is causing trouble, if one is available.
794 - Use a NIC whose driver does not have VLAN problems.
796 - Use "VLAN splinters", a feature in Open vSwitch 1.4 and later
797 that works around bugs in kernel drivers. To enable VLAN
798 splinters on interface eth0, use the command:
800 ovs-vsctl set interface eth0 other-config:enable-vlan-splinters=true
802 For VLAN splinters to be effective, Open vSwitch must know
803 which VLANs are in use. See the "VLAN splinters" section in
804 the Interface table in ovs-vswitchd.conf.db(5) for details on
805 how Open vSwitch infers in-use VLANs.
807 VLAN splinters increase memory use and reduce performance, so
808 use them only if needed.
810 - Apply the "vlan workaround" patch from the XenServer kernel
811 patch queue, build Open vSwitch against this patched kernel,
812 and then use ovs-vlan-bug-workaround(8) to enable the VLAN
813 workaround for each interface whose driver is buggy.
815 (This is a nontrivial exercise, so this option is included
816 only for completeness.)
818 It is not always easy to tell whether a Linux kernel driver has
819 buggy VLAN support. The ovs-vlan-test(8) and ovs-test(8) utilities
820 can help you test. See their manpages for details. Of the two
821 utilities, ovs-test(8) is newer and more thorough, but
822 ovs-vlan-test(8) may be easier to use.
824 Q: VLANs still don't work. I've tested the driver so I know that it's OK.
826 A: Do you have VLANs enabled on the physical switch that OVS is
827 attached to? Make sure that the port is configured to trunk the
828 VLAN or VLANs that you are using with OVS.
830 Q: Outgoing VLAN-tagged traffic goes through OVS to my physical switch
831 and to its destination host, but OVS seems to drop incoming return
834 A: It's possible that you have the VLAN configured on your physical
835 switch as the "native" VLAN. In this mode, the switch treats
836 incoming packets either tagged with the native VLAN or untagged as
837 part of the native VLAN. It may also send outgoing packets in the
838 native VLAN without a VLAN tag.
840 If this is the case, you have two choices:
842 - Change the physical switch port configuration to tag packets
843 it forwards to OVS with the native VLAN instead of forwarding
846 - Change the OVS configuration for the physical port to a
847 native VLAN mode. For example, the following sets up a
848 bridge with port eth0 in "native-tagged" mode in VLAN 9:
851 ovs-vsctl add-port br0 eth0 tag=9 vlan_mode=native-tagged
853 In this situation, "native-untagged" mode will probably work
854 equally well. Refer to the documentation for the Port table
855 in ovs-vswitchd.conf.db(5) for more information.
857 Q: I added a pair of VMs on different VLANs, like this:
860 ovs-vsctl add-port br0 eth0
861 ovs-vsctl add-port br0 tap0 tag=9
862 ovs-vsctl add-port br0 tap1 tag=10
864 but the VMs can't access each other, the external network, or the
867 A: It is to be expected that the VMs can't access each other. VLANs
868 are a means to partition a network. When you configured tap0 and
869 tap1 as access ports for different VLANs, you indicated that they
870 should be isolated from each other.
872 As for the external network and the Internet, it seems likely that
873 the machines you are trying to access are not on VLAN 9 (or 10) and
874 that the Internet is not available on VLAN 9 (or 10).
876 Q: I added a pair of VMs on the same VLAN, like this:
879 ovs-vsctl add-port br0 eth0
880 ovs-vsctl add-port br0 tap0 tag=9
881 ovs-vsctl add-port br0 tap1 tag=9
883 The VMs can access each other, but not the external network or the
886 A: It seems likely that the machines you are trying to access in the
887 external network are not on VLAN 9 and that the Internet is not
888 available on VLAN 9. Also, ensure VLAN 9 is set up as an allowed
889 trunk VLAN on the upstream switch port to which eth0 is connected.
891 Q: Can I configure an IP address on a VLAN?
893 A: Yes. Use an "internal port" configured as an access port. For
894 example, the following configures IP address 192.168.0.7 on VLAN 9.
895 That is, OVS will forward packets from eth0 to 192.168.0.7 only if
896 they have an 802.1Q header with VLAN 9. Conversely, traffic
897 forwarded from 192.168.0.7 to eth0 will be tagged with an 802.1Q
901 ovs-vsctl add-port br0 eth0
902 ovs-vsctl add-port br0 vlan9 tag=9 -- set interface vlan9 type=internal
903 ifconfig vlan9 192.168.0.7
905 Q: My OpenFlow controller doesn't see the VLANs that I expect.
907 A: The configuration for VLANs in the Open vSwitch database (e.g. via
908 ovs-vsctl) only affects traffic that goes through Open vSwitch's
909 implementation of the OpenFlow "normal switching" action. By
910 default, when Open vSwitch isn't connected to a controller and
911 nothing has been manually configured in the flow table, all traffic
912 goes through the "normal switching" action. But, if you set up
913 OpenFlow flows on your own, through a controller or using ovs-ofctl
914 or through other means, then you have to implement VLAN handling
917 You can use "normal switching" as a component of your OpenFlow
918 actions, e.g. by putting "normal" into the lists of actions on
919 ovs-ofctl or by outputting to OFPP_NORMAL from an OpenFlow
920 controller. In situations where this is not suitable, you can
921 implement VLAN handling yourself, e.g.:
923 - If a packet comes in on an access port, and the flow table
924 needs to send it out on a trunk port, then the flow can add
925 the appropriate VLAN tag with the "mod_vlan_vid" action.
927 - If a packet comes in on a trunk port, and the flow table
928 needs to send it out on an access port, then the flow can
929 strip the VLAN tag with the "strip_vlan" action.
931 Q: I configured ports on a bridge as access ports with different VLAN
935 ovs-vsctl set-controller br0 tcp:192.168.0.10:6633
936 ovs-vsctl add-port br0 eth0
937 ovs-vsctl add-port br0 tap0 tag=9
938 ovs-vsctl add-port br0 tap1 tag=10
940 but the VMs running behind tap0 and tap1 can still communicate,
941 that is, they are not isolated from each other even though they are
944 A: Do you have a controller configured on br0 (as the commands above
945 do)? If so, then this is a variant on the previous question, "My
946 OpenFlow controller doesn't see the VLANs that I expect," and you
947 can refer to the answer there for more information.
955 A: VXLAN stands for Virtual eXtensible Local Area Network, and is a means
956 to solve the scaling challenges of VLAN networks in a multi-tenant
957 environment. VXLAN is an overlay network which transports an L2 network
958 over an existing L3 network. For more information on VXLAN, please see
959 the IETF draft available here:
961 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03
963 Q: How much of the VXLAN protocol does Open vSwitch currently support?
965 A: Open vSwitch currently supports the framing format for packets on the
966 wire. There is currently no support for the multicast aspects of VXLAN.
967 To get around the lack of multicast support, it is possible to
968 pre-provision MAC to IP address mappings either manually or from a
971 Q: What destination UDP port does the VXLAN implementation in Open vSwitch
974 A: By default, Open vSwitch will use the assigned IANA port for VXLAN, which
975 is 4789. However, it is possible to configure the destination UDP port
976 manually on a per-VXLAN tunnel basis. An example of this configuration is
980 ovs-vsctl add-port br0 vxlan1 -- set interface vxlan1
981 type=vxlan options:remote_ip=192.168.1.2 options:key=flow
982 options:dst_port=8472
985 Using OpenFlow (Manually or Via Controller)
986 -------------------------------------------
988 Q: What versions of OpenFlow does Open vSwitch support?
990 A: Open vSwitch 1.9 and earlier support only OpenFlow 1.0 (plus
991 extensions that bring in many of the features from later versions
994 Open vSwitch 1.10 and later have experimental support for OpenFlow
995 1.2 and 1.3. On these versions of Open vSwitch, the following
996 command enables OpenFlow 1.0, 1.2, and 1.3 on bridge br0:
998 ovs-vsctl set bridge br0 protocols=OpenFlow10,OpenFlow12,OpenFlow13
1000 Open vSwitch version 1.12 and later will have experimental support
1001 for OpenFlow 1.1, 1.2, and 1.3. On these versions of Open vSwitch,
1002 the following command enables OpenFlow 1.0, 1.1, 1.2, and 1.3 on
1005 ovs-vsctl set bridge br0 protocols=OpenFlow10,OpenFlow11,OpenFlow12,OpenFlow13
1007 Use the -O option to enable support for later versions of OpenFlow
1008 in ovs-ofctl. For example:
1010 ovs-ofctl -O OpenFlow13 dump-flows br0
1012 Support for OpenFlow 1.1, 1.2, and 1.3 is still incomplete. Work
1013 to be done is tracked in OPENFLOW-1.1+ in the Open vSwitch sources
1014 (also via http://openvswitch.org/development/openflow-1-x-plan/).
1015 When support for a given OpenFlow version is solidly implemented,
1016 Open vSwitch will enable that version by default.
1018 Q: I'm getting "error type 45250 code 0". What's that?
1020 A: This is a Open vSwitch extension to OpenFlow error codes. Open
1021 vSwitch uses this extension when it must report an error to an
1022 OpenFlow controller but no standard OpenFlow error code is
1025 Open vSwitch logs the errors that it sends to controllers, so the
1026 easiest thing to do is probably to look at the ovs-vswitchd log to
1027 find out what the error was.
1029 If you want to dissect the extended error message yourself, the
1030 format is documented in include/openflow/nicira-ext.h in the Open
1031 vSwitch source distribution. The extended error codes are
1032 documented in lib/ofp-errors.h.
1034 Q1: Some of the traffic that I'd expect my OpenFlow controller to see
1035 doesn't actually appear through the OpenFlow connection, even
1036 though I know that it's going through.
1037 Q2: Some of the OpenFlow flows that my controller sets up don't seem
1038 to apply to certain traffic, especially traffic between OVS and
1039 the controller itself.
1041 A: By default, Open vSwitch assumes that OpenFlow controllers are
1042 connected "in-band", that is, that the controllers are actually
1043 part of the network that is being controlled. In in-band mode,
1044 Open vSwitch sets up special "hidden" flows to make sure that
1045 traffic can make it back and forth between OVS and the controllers.
1046 These hidden flows are higher priority than any flows that can be
1047 set up through OpenFlow, and they are not visible through normal
1048 OpenFlow flow table dumps.
1050 Usually, the hidden flows are desirable and helpful, but
1051 occasionally they can cause unexpected behavior. You can view the
1052 full OpenFlow flow table, including hidden flows, on bridge br0
1055 ovs-appctl bridge/dump-flows br0
1057 to help you debug. The hidden flows are those with priorities
1058 greater than 65535 (the maximum priority that can be set with
1061 The DESIGN file at the top level of the Open vSwitch source
1062 distribution describes the in-band model in detail.
1064 If your controllers are not actually in-band (e.g. they are on
1065 localhost via 127.0.0.1, or on a separate network), then you should
1066 configure your controllers in "out-of-band" mode. If you have one
1067 controller on bridge br0, then you can configure out-of-band mode
1070 ovs-vsctl set controller br0 connection-mode=out-of-band
1072 Q: I configured all my controllers for out-of-band control mode but
1073 "ovs-appctl bridge/dump-flows" still shows some hidden flows.
1075 A: You probably have a remote manager configured (e.g. with "ovs-vsctl
1076 set-manager"). By default, Open vSwitch assumes that managers need
1077 in-band rules set up on every bridge. You can disable these rules
1080 ovs-vsctl set bridge br0 other-config:disable-in-band=true
1082 This actually disables in-band control entirely for the bridge, as
1083 if all the bridge's controllers were configured for out-of-band
1086 Q: My OpenFlow controller doesn't see the VLANs that I expect.
1088 A: See answer under "VLANs", above.
1090 Q: I ran "ovs-ofctl add-flow br0 nw_dst=192.168.0.1,actions=drop"
1091 but I got a funny message like this:
1093 ofp_util|INFO|normalization changed ofp_match, details:
1094 ofp_util|INFO| pre: nw_dst=192.168.0.1
1097 and when I ran "ovs-ofctl dump-flows br0" I saw that my nw_dst
1098 match had disappeared, so that the flow ends up matching every
1101 A: The term "normalization" in the log message means that a flow
1102 cannot match on an L3 field without saying what L3 protocol is in
1103 use. The "ovs-ofctl" command above didn't specify an L3 protocol,
1104 so the L3 field match was dropped.
1106 In this case, the L3 protocol could be IP or ARP. A correct
1107 command for each possibility is, respectively:
1109 ovs-ofctl add-flow br0 ip,nw_dst=192.168.0.1,actions=drop
1113 ovs-ofctl add-flow br0 arp,nw_dst=192.168.0.1,actions=drop
1115 Similarly, a flow cannot match on an L4 field without saying what
1116 L4 protocol is in use. For example, the flow match "tp_src=1234"
1117 is, by itself, meaningless and will be ignored. Instead, to match
1118 TCP source port 1234, write "tcp,tp_src=1234", or to match UDP
1119 source port 1234, write "udp,tp_src=1234".
1121 Q: How can I figure out the OpenFlow port number for a given port?
1123 A: The OFPT_FEATURES_REQUEST message requests an OpenFlow switch to
1124 respond with an OFPT_FEATURES_REPLY that, among other information,
1125 includes a mapping between OpenFlow port names and numbers. From a
1126 command prompt, "ovs-ofctl show br0" makes such a request and
1127 prints the response for switch br0.
1129 The Interface table in the Open vSwitch database also maps OpenFlow
1130 port names to numbers. To print the OpenFlow port number
1131 associated with interface eth0, run:
1133 ovs-vsctl get Interface eth0 ofport
1135 You can print the entire mapping with:
1137 ovs-vsctl -- --columns=name,ofport list Interface
1139 but the output mixes together interfaces from all bridges in the
1140 database, so it may be confusing if more than one bridge exists.
1142 In the Open vSwitch database, ofport value -1 means that the
1143 interface could not be created due to an error. (The Open vSwitch
1144 log should indicate the reason.) ofport value [] (the empty set)
1145 means that the interface hasn't been created yet. The latter is
1146 normally an intermittent condition (unless ovs-vswitchd is not
1149 Q: I added some flows with my controller or with ovs-ofctl, but when I
1150 run "ovs-dpctl dump-flows" I don't see them.
1152 A: ovs-dpctl queries a kernel datapath, not an OpenFlow switch. It
1153 won't display the information that you want. You want to use
1154 "ovs-ofctl dump-flows" instead.
1156 Q: It looks like each of the interfaces in my bonded port shows up
1157 as an individual OpenFlow port. Is that right?
1159 A: Yes, Open vSwitch makes individual bond interfaces visible as
1160 OpenFlow ports, rather than the bond as a whole. The interfaces
1161 are treated together as a bond for only a few purposes:
1163 - Sending a packet to the OFPP_NORMAL port. (When an OpenFlow
1164 controller is not configured, this happens implicitly to
1167 - Mirrors configured for output to a bonded port.
1169 It would make a lot of sense for Open vSwitch to present a bond as
1170 a single OpenFlow port. If you want to contribute an
1171 implementation of such a feature, please bring it up on the Open
1172 vSwitch development mailing list at dev@openvswitch.org.
1174 Q: I have a sophisticated network setup involving Open vSwitch, VMs or
1175 multiple hosts, and other components. The behavior isn't what I
1178 A: To debug network behavior problems, trace the path of a packet,
1179 hop-by-hop, from its origin in one host to a remote host. If
1180 that's correct, then trace the path of the response packet back to
1183 Usually a simple ICMP echo request and reply ("ping") packet is
1184 good enough. Start by initiating an ongoing "ping" from the origin
1185 host to a remote host. If you are tracking down a connectivity
1186 problem, the "ping" will not display any successful output, but
1187 packets are still being sent. (In this case the packets being sent
1188 are likely ARP rather than ICMP.)
1190 Tools available for tracing include the following:
1192 - "tcpdump" and "wireshark" for observing hops across network
1193 devices, such as Open vSwitch internal devices and physical
1196 - "ovs-appctl dpif/dump-flows <br>" in Open vSwitch 1.10 and
1197 later or "ovs-dpctl dump-flows <br>" in earlier versions.
1198 These tools allow one to observe the actions being taken on
1199 packets in ongoing flows.
1201 See ovs-vswitchd(8) for "ovs-appctl dpif/dump-flows"
1202 documentation, ovs-dpctl(8) for "ovs-dpctl dump-flows"
1203 documentation, and "Why are there so many different ways to
1204 dump flows?" above for some background.
1206 - "ovs-appctl ofproto/trace" to observe the logic behind how
1207 ovs-vswitchd treats packets. See ovs-vswitchd(8) for
1208 documentation. You can out more details about a given flow
1209 that "ovs-dpctl dump-flows" displays, by cutting and pasting
1210 a flow from the output into an "ovs-appctl ofproto/trace"
1213 - SPAN, RSPAN, and ERSPAN features of physical switches, to
1214 observe what goes on at these physical hops.
1216 Starting at the origin of a given packet, observe the packet at
1217 each hop in turn. For example, in one plausible scenario, you
1220 1. "tcpdump" the "eth" interface through which an ARP egresses
1221 a VM, from inside the VM.
1223 2. "tcpdump" the "vif" or "tap" interface through which the ARP
1224 ingresses the host machine.
1226 3. Use "ovs-dpctl dump-flows" to spot the ARP flow and observe
1227 the host interface through which the ARP egresses the
1228 physical machine. You may need to use "ovs-dpctl show" to
1229 interpret the port numbers. If the output seems surprising,
1230 you can use "ovs-appctl ofproto/trace" to observe details of
1231 how ovs-vswitchd determined the actions in the "ovs-dpctl
1234 4. "tcpdump" the "eth" interface through which the ARP egresses
1235 the physical machine.
1237 5. "tcpdump" the "eth" interface through which the ARP
1238 ingresses the physical machine, at the remote host that
1241 6. Use "ovs-dpctl dump-flows" to spot the ARP flow on the
1242 remote host that receives the ARP and observe the VM "vif"
1243 or "tap" interface to which the flow is directed. Again,
1244 "ovs-dpctl show" and "ovs-appctl ofproto/trace" might help.
1246 7. "tcpdump" the "vif" or "tap" interface to which the ARP is
1249 8. "tcpdump" the "eth" interface through which the ARP
1250 ingresses a VM, from inside the VM.
1252 It is likely that during one of these steps you will figure out the
1253 problem. If not, then follow the ARP reply back to the origin, in
1256 Q: How do I make a flow drop packets?
1258 A: An empty set of actions causes a packet to be dropped. You can
1259 specify an empty set of actions with "actions=" on the ovs-ofctl
1260 command line. For example:
1262 ovs-ofctl add-flow br0 priority=65535,actions=
1264 would cause every packet entering switch br0 to be dropped.
1266 You can write "drop" explicitly if you like. The effect is the
1267 same. Thus, the following command also causes every packet
1268 entering switch br0 to be dropped:
1270 ovs-ofctl add-flow br0 priority=65535,actions=drop
1272 Q: I added a flow to send packets out the ingress port, like this:
1274 ovs-ofctl add-flow br0 in_port=2,actions=2
1276 but OVS drops the packets instead.
1278 A: Yes, OpenFlow requires a switch to ignore attempts to send a packet
1279 out its ingress port. The rationale is that dropping these packets
1280 makes it harder to loop the network. Sometimes this behavior can
1281 even be convenient, e.g. it is often the desired behavior in a flow
1282 that forwards a packet to several ports ("floods" the packet).
1284 Sometimes one really needs to send a packet out its ingress port.
1285 In this case, output to OFPP_IN_PORT, which in ovs-ofctl syntax is
1286 expressed as just "in_port", e.g.:
1288 ovs-ofctl add-flow br0 in_port=2,actions=in_port
1290 This also works in some circumstances where the flow doesn't match
1291 on the input port. For example, if you know that your switch has
1292 five ports numbered 2 through 6, then the following will send every
1293 received packet out every port, even its ingress port:
1295 ovs-ofctl add-flow br0 actions=2,3,4,5,6,in_port
1299 ovs-ofctl add-flow br0 actions=all,in_port
1301 Sometimes, in complicated flow tables with multiple levels of
1302 "resubmit" actions, a flow needs to output to a particular port
1303 that may or may not be the ingress port. It's difficult to take
1304 advantage of OFPP_IN_PORT in this situation. To help, Open vSwitch
1305 provides, as an OpenFlow extension, the ability to modify the
1306 in_port field. Whatever value is currently in the in_port field is
1307 the port to which outputs will be dropped, as well as the
1308 destination for OFPP_IN_PORT. This means that the following will
1309 reliably output to port 2 or to ports 2 through 6, respectively:
1311 ovs-ofctl add-flow br0 in_port=2,actions=load:0->NXM_OF_IN_PORT[],2
1312 ovs-ofctl add-flow br0 actions=load:0->NXM_OF_IN_PORT[],2,3,4,5,6
1314 If the input port is important, then one may save and restore it on
1317 ovs-ofctl add-flow br0 actions=push:NXM_OF_IN_PORT[],\
1318 load:0->NXM_OF_IN_PORT[],\
1320 pop:NXM_OF_IN_PORT[]
1322 Q: My bridge br0 has host 192.168.0.1 on port 1 and host 192.168.0.2
1323 on port 2. I set up flows to forward only traffic destined to the
1324 other host and drop other traffic, like this:
1326 priority=5,in_port=1,ip,nw_dst=192.168.0.2,actions=2
1327 priority=5,in_port=2,ip,nw_dst=192.168.0.1,actions=1
1328 priority=0,actions=drop
1330 But it doesn't work--I don't get any connectivity when I do this.
1333 A: These flows drop the ARP packets that IP hosts use to establish IP
1334 connectivity over Ethernet. To solve the problem, add flows to
1335 allow ARP to pass between the hosts:
1337 priority=5,in_port=1,arp,actions=2
1338 priority=5,in_port=2,arp,actions=1
1340 This issue can manifest other ways, too. The following flows that
1341 match on Ethernet addresses instead of IP addresses will also drop
1342 ARP packets, because ARP requests are broadcast instead of being
1343 directed to a specific host:
1345 priority=5,in_port=1,dl_dst=54:00:00:00:00:02,actions=2
1346 priority=5,in_port=2,dl_dst=54:00:00:00:00:01,actions=1
1347 priority=0,actions=drop
1349 The solution already described above will also work in this case.
1350 It may be better to add flows to allow all multicast and broadcast
1353 priority=5,in_port=1,dl_dst=01:00:00:00:00:00/01:00:00:00:00:00,actions=2
1354 priority=5,in_port=2,dl_dst=01:00:00:00:00:00/01:00:00:00:00:00,actions=1
1359 bugs@openvswitch.org
1360 http://openvswitch.org/
projects / mirror_ovs.git / blob