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 environments. A
13 vswitch forwards traffic between different VMs on the same physical host
14 and also forwards traffic between VMs and the physical network. Open
15 vSwitch supports standard management interfaces (e.g. sFlow, NetFlow,
16 RSPAN, CLI), and is open to programmatic extension and control using
17 OpenFlow and the OVSDB management protocol.
19 Open vSwitch as designed to be compatible with modern switching
20 chipsets. This means that it can be ported to existing high-fanout
21 switches allowing the same flexible control of the physical
22 infrastructure as the virtual infrastructure. It also means that
23 Open vSwitch will be able to take advantage of on-NIC switching
24 chipsets as their functionality matures.
26 Q: What virtualization platforms can use Open vSwitch?
28 A: Open vSwitch can currently run on any Linux-based virtualization
29 platform (kernel 2.6.18 and newer), including: KVM, VirtualBox, Xen,
30 Xen Cloud Platform, XenServer. As of Linux 3.3 it is part of the
31 mainline kernel. The bulk of the code is written in platform-
32 independent C and is easily ported to other environments. We welcome
33 inquires about integrating Open vSwitch with other virtualization
36 Q: How can I try Open vSwitch?
38 A: The Open vSwitch source code can be built on a Linux system. You can
39 build and experiment with Open vSwitch on any Linux machine.
40 Packages for various Linux distributions are available on many
41 platforms, including: Debian, Ubuntu, Fedora.
43 You may also download and run a virtualization platform that already
44 has Open vSwitch integrated. For example, download a recent ISO for
45 XenServer or Xen Cloud Platform. Be aware that the version
46 integrated with a particular platform may not be the most recent Open
49 Q: Does Open vSwitch only work on Linux?
51 A: No, Open vSwitch has been ported to a number of different operating
52 systems and hardware platforms. Most of the development work occurs
53 on Linux, but the code should be portable to any POSIX system. We've
54 seen Open vSwitch ported to a number of different platforms,
55 including FreeBSD, Windows, and even non-POSIX embedded systems.
57 By definition, the Open vSwitch Linux kernel module only works on
58 Linux and will provide the highest performance. However, a userspace
59 datapath is available that should be very portable.
61 Q: What's involved with porting Open vSwitch to a new platform or
64 A: The PORTING document describes how one would go about porting Open
65 vSwitch to a new operating system or hardware platform.
67 Q: Why would I use Open vSwitch instead of the Linux bridge?
69 A: Open vSwitch is specially designed to make it easier to manage VM
70 network configuration and monitor state spread across many physical
71 hosts in dynamic virtualized environments. Please see WHY-OVS for a
72 more detailed description of how Open vSwitch relates to the Linux
75 Q: How is Open vSwitch related to distributed virtual switches like the
76 VMware vNetwork distributed switch or the Cisco Nexus 1000V?
78 A: Distributed vswitch applications (e.g., VMware vNetwork distributed
79 switch, Cisco Nexus 1000V) provide a centralized way to configure and
80 monitor the network state of VMs that are spread across many physical
81 hosts. Open vSwitch is not a distributed vswitch itself, rather it
82 runs on each physical host and supports remote management in a way
83 that makes it easier for developers of virtualization/cloud
84 management platforms to offer distributed vswitch capabilities.
86 To aid in distribution, Open vSwitch provides two open protocols that
87 are specially designed for remote management in virtualized network
88 environments: OpenFlow, which exposes flow-based forwarding state,
89 and the OVSDB management protocol, which exposes switch port state.
90 In addition to the switch implementation itself, Open vSwitch
91 includes tools (ovs-controller, ovs-ofctl, ovs-vsctl) that developers
92 can script and extend to provide distributed vswitch capabilities
93 that are closely integrated with their virtualization management
96 Q: Why doesn't Open vSwitch support distribution?
98 A: Open vSwitch is intended to be a useful component for building
99 flexible network infrastructure. There are many different approaches
100 to distribution which balance trade-offs between simplicity,
101 scalability, hardware compatibility, convergence times, logical
102 forwarding model, etc. The goal of Open vSwitch is to be able to
103 support all as a primitive building block rather than choose a
104 particular point in the distributed design space.
106 Q: How can I contribute to the Open vSwitch Community?
108 A: You can start by joining the mailing lists and helping to answer
109 questions. You can also suggest improvements to documentation. If
110 you have a feature or bug you would like to work on, send a mail to
111 one of the mailing lists:
113 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 ------------ -------------
151 Open vSwitch userspace should also work with the Linux kernel module
152 built into Linux 3.3 and later.
154 Open vSwitch userspace is not sensitive to the Linux kernel version.
155 It should build against almost any kernel, certainly against 2.6.18
158 Q: Should userspace or kernel be upgraded first to minimize downtime?
160 In general, the Open vSwitch userspace should be used with the
161 kernel version included in the same release or with the version
162 from upstream Linux. However, when upgrading between two releases
163 of Open vSwitch it is best to migrate userspace first to reduce
164 the possbility of incompatibilities.
166 Q: What features are not available in the Open vSwitch kernel datapath
167 that ships as part of the upstream Linux kernel?
169 A: The kernel module in upstream Linux 3.3 and later does not include
170 tunnel virtual ports, that is, interfaces with type "gre",
171 "ipsec_gre", "gre64", "ipsec_gre64", or "vxlan". It is
172 possible to create tunnels in Linux and attach them to Open vSwitch
173 as system devices. However, they cannot be dynamically created
174 through the OVSDB protocol or set the tunnel ids as a flow action.
176 Work is in progress in adding tunnel virtual ports to the upstream
177 Linux version of the Open vSwitch kernel module. For now, if you
178 need these features, use the kernel module from the Open vSwitch
179 distribution instead of the upstream Linux kernel module.
181 The upstream kernel module does not include patch ports, but this
182 only matters for Open vSwitch 1.9 and earlier, because Open vSwitch
183 1.10 and later implement patch ports without using this kernel
186 Q: What features are not available when using the userspace datapath?
188 A: Tunnel virtual ports are not supported, as described in the
189 previous answer. It is also not possible to use queue-related
190 actions. On Linux kernels before 2.6.39, maximum-sized VLAN packets
191 may not be transmitted.
197 Q: I thought Open vSwitch was a virtual Ethernet switch, but the
198 documentation keeps talking about bridges. What's a bridge?
200 A: In networking, the terms "bridge" and "switch" are synonyms. Open
201 vSwitch implements an Ethernet switch, which means that it is also
206 A: See the "VLAN" section below.
212 Q: How do I configure a port as an access port?
214 A: Add "tag=VLAN" to your "ovs-vsctl add-port" command. For example,
215 the following commands configure br0 with eth0 as a trunk port (the
216 default) and tap0 as an access port for VLAN 9:
219 ovs-vsctl add-port br0 eth0
220 ovs-vsctl add-port br0 tap0 tag=9
222 If you want to configure an already added port as an access port,
223 use "ovs-vsctl set", e.g.:
225 ovs-vsctl set port tap0 tag=9
227 Q: How do I configure a port as a SPAN port, that is, enable mirroring
228 of all traffic to that port?
230 A: The following commands configure br0 with eth0 and tap0 as trunk
231 ports. All traffic coming in or going out on eth0 or tap0 is also
232 mirrored to tap1; any traffic arriving on tap1 is dropped:
235 ovs-vsctl add-port br0 eth0
236 ovs-vsctl add-port br0 tap0
237 ovs-vsctl add-port br0 tap1 \
238 -- --id=@p get port tap1 \
239 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
240 -- set bridge br0 mirrors=@m
242 To later disable mirroring, run:
244 ovs-vsctl clear bridge br0 mirrors
246 Q: How do I configure a VLAN as an RSPAN VLAN, that is, enable
247 mirroring of all traffic to that VLAN?
249 A: The following commands configure br0 with eth0 as a trunk port and
250 tap0 as an access port for VLAN 10. All traffic coming in or going
251 out on tap0, as well as traffic coming in or going out on eth0 in
252 VLAN 10, is also mirrored to VLAN 15 on eth0. The original tag for
253 VLAN 10, in cases where one is present, is dropped as part of
257 ovs-vsctl add-port br0 eth0
258 ovs-vsctl add-port br0 tap0 tag=10
260 -- --id=@m create mirror name=m0 select-all=true select-vlan=10 \
262 -- set bridge br0 mirrors=@m
264 To later disable mirroring, run:
266 ovs-vsctl clear bridge br0 mirrors
268 Mirroring to a VLAN can disrupt a network that contains unmanaged
269 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
270 GRE tunnel has fewer caveats than mirroring to a VLAN and should
271 generally be preferred.
273 Q: Can I mirror more than one input VLAN to an RSPAN VLAN?
275 A: Yes, but mirroring to a VLAN strips the original VLAN tag in favor
276 of the specified output-vlan. This loss of information may make
277 the mirrored traffic too hard to interpret.
279 To mirror multiple VLANs, use the commands above, but specify a
280 comma-separated list of VLANs as the value for select-vlan. To
281 mirror every VLAN, use the commands above, but omit select-vlan and
284 When a packet arrives on a VLAN that is used as a mirror output
285 VLAN, the mirror is disregarded. Instead, in standalone mode, OVS
286 floods the packet across all the ports for which the mirror output
287 VLAN is configured. (If an OpenFlow controller is in use, then it
288 can override this behavior through the flow table.) If OVS is used
289 as an intermediate switch, rather than an edge switch, this ensures
290 that the RSPAN traffic is distributed through the network.
292 Mirroring to a VLAN can disrupt a network that contains unmanaged
293 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
294 GRE tunnel has fewer caveats than mirroring to a VLAN and should
295 generally be preferred.
297 Q: How do I configure mirroring of all traffic to a GRE tunnel?
299 A: The following commands configure br0 with eth0 and tap0 as trunk
300 ports. All traffic coming in or going out on eth0 or tap0 is also
301 mirrored to gre0, a GRE tunnel to the remote host 192.168.1.10; any
302 traffic arriving on gre0 is dropped:
305 ovs-vsctl add-port br0 eth0
306 ovs-vsctl add-port br0 tap0
307 ovs-vsctl add-port br0 gre0 \
308 -- set interface gre0 type=gre options:remote_ip=192.168.1.10 \
309 -- --id=@p get port gre0 \
310 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
311 -- set bridge br0 mirrors=@m
313 To later disable mirroring and destroy the GRE tunnel:
315 ovs-vsctl clear bridge br0 mirrors
316 ovs-vcstl del-port br0 gre0
318 Q: Does Open vSwitch support ERSPAN?
320 A: No. ERSPAN is an undocumented proprietary protocol. As an
321 alternative, Open vSwitch supports mirroring to a GRE tunnel (see
324 Q: Why are there so many different ways to dump flows?
326 A: Open vSwitch uses different kinds of flows for different purposes:
328 - OpenFlow flows are the most important kind of flow. OpenFlow
329 controllers use these flows to define a switch's policy.
330 OpenFlow flows support wildcards, priorities, and multiple
333 When in-band control is in use, Open vSwitch sets up a few
334 "hidden" flows, with priority higher than a controller or the
335 user can configure, that are not visible via OpenFlow. (See
336 the "Controller" section of the FAQ for more information
339 - The Open vSwitch software switch implementation uses a second
340 kind of flow internally. These flows, called "exact-match"
341 or "datapath" or "kernel" flows, do not support wildcards or
342 priorities and comprise only a single table, which makes them
343 suitable for caching. OpenFlow flows and exact-match flows
344 also support different actions and number ports differently.
346 Exact-match flows are an implementation detail that is
347 subject to change in future versions of Open vSwitch. Even
348 with the current version of Open vSwitch, hardware switch
349 implementations do not necessarily use exact-match flows.
351 Each of the commands for dumping flows has a different purpose:
353 - "ovs-ofctl dump-flows <br>" dumps OpenFlow flows, excluding
354 hidden flows. This is the most commonly useful form of flow
355 dump. (Unlike the other commands, this should work with any
356 OpenFlow switch, not just Open vSwitch.)
358 - "ovs-appctl bridge/dump-flows <br>" dumps OpenFlow flows,
359 including hidden flows. This is occasionally useful for
360 troubleshooting suspected issues with in-band control.
362 - "ovs-dpctl dump-flows [dp]" dumps the exact-match flow table
363 entries for a Linux kernel-based datapath. In Open vSwitch
364 1.10 and later, ovs-vswitchd later merges multiple switches
365 into a single datapath, so it will show all the flows on all
366 your kernel-based switches. This command can occasionally be
367 useful for debugging.
369 - "ovs-appctl dpif/dump-flows <br>", new in Open vSwitch 1.10,
370 dumps exact-match flows for only the specified bridge,
371 regardless of the type.
374 Configuration Problems
375 ----------------------
377 Q: I created a bridge and added my Ethernet port to it, using commands
381 ovs-vsctl add-port br0 eth0
383 and as soon as I ran the "add-port" command I lost all connectivity
386 A: A physical Ethernet device that is part of an Open vSwitch bridge
387 should not have an IP address. If one does, then that IP address
388 will not be fully functional.
390 You can restore functionality by moving the IP address to an Open
391 vSwitch "internal" device, such as the network device named after
392 the bridge itself. For example, assuming that eth0's IP address is
393 192.168.128.5, you could run the commands below to fix up the
396 ifconfig eth0 0.0.0.0
397 ifconfig br0 192.168.128.5
399 (If your only connection to the machine running OVS is through the
400 IP address in question, then you would want to run all of these
401 commands on a single command line, or put them into a script.) If
402 there were any additional routes assigned to eth0, then you would
403 also want to use commands to adjust these routes to go through br0.
405 If you use DHCP to obtain an IP address, then you should kill the
406 DHCP client that was listening on the physical Ethernet interface
407 (e.g. eth0) and start one listening on the internal interface
408 (e.g. br0). You might still need to manually clear the IP address
409 from the physical interface (e.g. with "ifconfig eth0 0.0.0.0").
411 There is no compelling reason why Open vSwitch must work this way.
412 However, this is the way that the Linux kernel bridge module has
413 always worked, so it's a model that those accustomed to Linux
414 bridging are already used to. Also, the model that most people
415 expect is not implementable without kernel changes on all the
416 versions of Linux that Open vSwitch supports.
418 By the way, this issue is not specific to physical Ethernet
419 devices. It applies to all network devices except Open vswitch
422 Q: I created a bridge and added a couple of Ethernet ports to it,
423 using commands like these:
426 ovs-vsctl add-port br0 eth0
427 ovs-vsctl add-port br0 eth1
429 and now my network seems to have melted: connectivity is unreliable
430 (even connectivity that doesn't go through Open vSwitch), all the
431 LEDs on my physical switches are blinking, wireshark shows
432 duplicated packets, and CPU usage is very high.
434 A: More than likely, you've looped your network. Probably, eth0 and
435 eth1 are connected to the same physical Ethernet switch. This
436 yields a scenario where OVS receives a broadcast packet on eth0 and
437 sends it out on eth1, then the physical switch connected to eth1
438 sends the packet back on eth0, and so on forever. More complicated
439 scenarios, involving a loop through multiple switches, are possible
442 The solution depends on what you are trying to do:
444 - If you added eth0 and eth1 to get higher bandwidth or higher
445 reliability between OVS and your physical Ethernet switch,
446 use a bond. The following commands create br0 and then add
447 eth0 and eth1 as a bond:
450 ovs-vsctl add-bond br0 bond0 eth0 eth1
452 Bonds have tons of configuration options. Please read the
453 documentation on the Port table in ovs-vswitchd.conf.db(5)
456 - Perhaps you don't actually need eth0 and eth1 to be on the
457 same bridge. For example, if you simply want to be able to
458 connect each of them to virtual machines, then you can put
459 each of them on a bridge of its own:
462 ovs-vsctl add-port br0 eth0
465 ovs-vsctl add-port br1 eth1
467 and then connect VMs to br0 and br1. (A potential
468 disadvantage is that traffic cannot directly pass between br0
469 and br1. Instead, it will go out eth0 and come back in eth1,
472 - If you have a redundant or complex network topology and you
473 want to prevent loops, turn on spanning tree protocol (STP).
474 The following commands create br0, enable STP, and add eth0
475 and eth1 to the bridge. The order is important because you
476 don't want have to have a loop in your network even
480 ovs-vsctl set bridge br0 stp_enable=true
481 ovs-vsctl add-port br0 eth0
482 ovs-vsctl add-port br0 eth1
484 The Open vSwitch implementation of STP is not well tested.
485 Please report any bugs you observe, but if you'd rather avoid
486 acting as a beta tester then another option might be your
489 Q: I can't seem to use Open vSwitch in a wireless network.
491 A: Wireless base stations generally only allow packets with the source
492 MAC address of NIC that completed the initial handshake.
493 Therefore, without MAC rewriting, only a single device can
494 communicate over a single wireless link.
496 This isn't specific to Open vSwitch, it's enforced by the access
497 point, so the same problems will show up with the Linux bridge or
498 any other way to do bridging.
500 Q: I can't seem to add my PPP interface to an Open vSwitch bridge.
502 A: PPP most commonly carries IP packets, but Open vSwitch works only
503 with Ethernet frames. The correct way to interface PPP to an
504 Ethernet network is usually to use routing instead of switching.
506 Q: Is there any documentation on the database tables and fields?
508 A: Yes. ovs-vswitchd.conf.db(5) is a comprehensive reference.
510 Q: When I run ovs-dpctl I no longer see the bridges I created. Instead,
511 I only see a datapath called "ovs-system". How can I see datapath
512 information about a particular bridge?
514 A: In version 1.9.0, OVS switched to using a single datapath that is
515 shared by all bridges of that type. The "ovs-appctl dpif/*"
516 commands provide similar functionality that is scoped by the bridge.
519 Quality of Service (QoS)
520 ------------------------
522 Q: How do I configure Quality of Service (QoS)?
524 A: Suppose that you want to set up bridge br0 connected to physical
525 Ethernet port eth0 (a 1 Gbps device) and virtual machine interfaces
526 vif1.0 and vif2.0, and that you want to limit traffic from vif1.0
527 to eth0 to 10 Mbps and from vif2.0 to eth0 to 20 Mbps. Then, you
528 could configure the bridge this way:
532 add-port br0 eth0 -- \
533 add-port br0 vif1.0 -- set interface vif1.0 ofport_request=5 -- \
534 add-port br0 vif2.0 -- set interface vif2.0 ofport_request=6 -- \
535 set port eth0 qos=@newqos -- \
536 --id=@newqos create qos type=linux-htb \
537 other-config:max-rate=1000000000 \
538 queues:123=@vif10queue \
539 queues:234=@vif20queue -- \
540 --id=@vif10queue create queue other-config:max-rate=10000000 -- \
541 --id=@vif20queue create queue other-config:max-rate=20000000
543 At this point, bridge br0 is configured with the ports and eth0 is
544 configured with the queues that you need for QoS, but nothing is
545 actually directing packets from vif1.0 or vif2.0 to the queues that
546 we have set up for them. That means that all of the packets to
547 eth0 are going to the "default queue", which is not what we want.
549 We use OpenFlow to direct packets from vif1.0 and vif2.0 to the
550 queues reserved for them:
552 ovs-ofctl add-flow br0 in_port=5,actions=set_queue:123,normal
553 ovs-ofctl add-flow br0 in_port=6,actions=set_queue:234,normal
555 Each of the above flows matches on the input port, sets up the
556 appropriate queue (123 for vif1.0, 234 for vif2.0), and then
557 executes the "normal" action, which performs the same switching
558 that Open vSwitch would have done without any OpenFlow flows being
559 present. (We know that vif1.0 and vif2.0 have OpenFlow port
560 numbers 5 and 6, respectively, because we set their ofport_request
561 columns above. If we had not done that, then we would have needed
562 to find out their port numbers before setting up these flows.)
564 Now traffic going from vif1.0 or vif2.0 to eth0 should be
567 By the way, if you delete the bridge created by the above commands,
572 then that will leave one unreferenced QoS record and two
573 unreferenced Queue records in the Open vSwich database. One way to
574 clear them out, assuming you don't have other QoS or Queue records
575 that you want to keep, is:
577 ovs-vsctl -- --all destroy QoS -- --all destroy Queue
579 Q: I configured Quality of Service (QoS) in my OpenFlow network by
580 adding records to the QoS and Queue table, but the results aren't
583 A: Did you install OpenFlow flows that use your queues? This is the
584 primary way to tell Open vSwitch which queues you want to use. If
585 you don't do this, then the default queue will be used, which will
586 probably not have the effect you want.
588 Refer to the previous question for an example.
590 Q: I configured QoS, correctly, but my measurements show that it isn't
591 working as well as I expect.
593 A: With the Linux kernel, the Open vSwitch implementation of QoS has
596 - Open vSwitch configures a subset of Linux kernel QoS
597 features, according to what is in OVSDB. It is possible that
598 this code has bugs. If you believe that this is so, then you
599 can configure the Linux traffic control (QoS) stack directly
600 with the "tc" program. If you get better results that way,
601 you can send a detailed bug report to bugs@openvswitch.org.
603 It is certain that Open vSwitch cannot configure every Linux
604 kernel QoS feature. If you need some feature that OVS cannot
605 configure, then you can also use "tc" directly (or add that
608 - The Open vSwitch implementation of OpenFlow allows flows to
609 be directed to particular queues. This is pretty simple and
610 unlikely to have serious bugs at this point.
612 However, most problems with QoS on Linux are not bugs in Open
613 vSwitch at all. They tend to be either configuration errors
614 (please see the earlier questions in this section) or issues with
615 the traffic control (QoS) stack in Linux. The Open vSwitch
616 developers are not experts on Linux traffic control. We suggest
617 that, if you believe you are encountering a problem with Linux
618 traffic control, that you consult the tc manpages (e.g. tc(8),
619 tc-htb(8), tc-hfsc(8)), web resources (e.g. http://lartc.org/), or
620 mailing lists (e.g. http://vger.kernel.org/vger-lists.html#netdev).
628 A: At the simplest level, a VLAN (short for "virtual LAN") is a way to
629 partition a single switch into multiple switches. Suppose, for
630 example, that you have two groups of machines, group A and group B.
631 You want the machines in group A to be able to talk to each other,
632 and you want the machine in group B to be able to talk to each
633 other, but you don't want the machines in group A to be able to
634 talk to the machines in group B. You can do this with two
635 switches, by plugging the machines in group A into one switch and
636 the machines in group B into the other switch.
638 If you only have one switch, then you can use VLANs to do the same
639 thing, by configuring the ports for machines in group A as VLAN
640 "access ports" for one VLAN and the ports for group B as "access
641 ports" for a different VLAN. The switch will only forward packets
642 between ports that are assigned to the same VLAN, so this
643 effectively subdivides your single switch into two independent
644 switches, one for each group of machines.
646 So far we haven't said anything about VLAN headers. With access
647 ports, like we've described so far, no VLAN header is present in
648 the Ethernet frame. This means that the machines (or switches)
649 connected to access ports need not be aware that VLANs are
650 involved, just like in the case where we use two different physical
653 Now suppose that you have a whole bunch of switches in your
654 network, instead of just one, and that some machines in group A are
655 connected directly to both switches 1 and 2. To allow these
656 machines to talk to each other, you could add an access port for
657 group A's VLAN to switch 1 and another to switch 2, and then
658 connect an Ethernet cable between those ports. That works fine,
659 but it doesn't scale well as the number of switches and the number
660 of VLANs increases, because you use up a lot of valuable switch
661 ports just connecting together your VLANs.
663 This is where VLAN headers come in. Instead of using one cable and
664 two ports per VLAN to connect a pair of switches, we configure a
665 port on each switch as a VLAN "trunk port". Packets sent and
666 received on a trunk port carry a VLAN header that says what VLAN
667 the packet belongs to, so that only two ports total are required to
668 connect the switches, regardless of the number of VLANs in use.
669 Normally, only switches (either physical or virtual) are connected
670 to a trunk port, not individual hosts, because individual hosts
671 don't expect to see a VLAN header in the traffic that they receive.
673 None of the above discussion says anything about particular VLAN
674 numbers. This is because VLAN numbers are completely arbitrary.
675 One must only ensure that a given VLAN is numbered consistently
676 throughout a network and that different VLANs are given different
677 numbers. (That said, VLAN 0 is usually synonymous with a packet
678 that has no VLAN header, and VLAN 4095 is reserved.)
682 A: Many drivers in Linux kernels before version 3.3 had VLAN-related
683 bugs. If you are having problems with VLANs that you suspect to be
684 driver related, then you have several options:
686 - Upgrade to Linux 3.3 or later.
688 - Build and install a fixed version of the particular driver
689 that is causing trouble, if one is available.
691 - Use a NIC whose driver does not have VLAN problems.
693 - Use "VLAN splinters", a feature in Open vSwitch 1.4 and later
694 that works around bugs in kernel drivers. To enable VLAN
695 splinters on interface eth0, use the command:
697 ovs-vsctl set interface eth0 other-config:enable-vlan-splinters=true
699 For VLAN splinters to be effective, Open vSwitch must know
700 which VLANs are in use. See the "VLAN splinters" section in
701 the Interface table in ovs-vswitchd.conf.db(5) for details on
702 how Open vSwitch infers in-use VLANs.
704 VLAN splinters increase memory use and reduce performance, so
705 use them only if needed.
707 - Apply the "vlan workaround" patch from the XenServer kernel
708 patch queue, build Open vSwitch against this patched kernel,
709 and then use ovs-vlan-bug-workaround(8) to enable the VLAN
710 workaround for each interface whose driver is buggy.
712 (This is a nontrivial exercise, so this option is included
713 only for completeness.)
715 It is not always easy to tell whether a Linux kernel driver has
716 buggy VLAN support. The ovs-vlan-test(8) and ovs-test(8) utilities
717 can help you test. See their manpages for details. Of the two
718 utilities, ovs-test(8) is newer and more thorough, but
719 ovs-vlan-test(8) may be easier to use.
721 Q: VLANs still don't work. I've tested the driver so I know that it's OK.
723 A: Do you have VLANs enabled on the physical switch that OVS is
724 attached to? Make sure that the port is configured to trunk the
725 VLAN or VLANs that you are using with OVS.
727 Q: Outgoing VLAN-tagged traffic goes through OVS to my physical switch
728 and to its destination host, but OVS seems to drop incoming return
731 A: It's possible that you have the VLAN configured on your physical
732 switch as the "native" VLAN. In this mode, the switch treats
733 incoming packets either tagged with the native VLAN or untagged as
734 part of the native VLAN. It may also send outgoing packets in the
735 native VLAN without a VLAN tag.
737 If this is the case, you have two choices:
739 - Change the physical switch port configuration to tag packets
740 it forwards to OVS with the native VLAN instead of forwarding
743 - Change the OVS configuration for the physical port to a
744 native VLAN mode. For example, the following sets up a
745 bridge with port eth0 in "native-tagged" mode in VLAN 9:
748 ovs-vsctl add-port br0 eth0 tag=9 vlan_mode=native-tagged
750 In this situation, "native-untagged" mode will probably work
751 equally well. Refer to the documentation for the Port table
752 in ovs-vswitchd.conf.db(5) for more information.
754 Q: I added a pair of VMs on different VLANs, like this:
757 ovs-vsctl add-port br0 eth0
758 ovs-vsctl add-port br0 tap0 tag=9
759 ovs-vsctl add-port br0 tap1 tag=10
761 but the VMs can't access each other, the external network, or the
764 A: It is to be expected that the VMs can't access each other. VLANs
765 are a means to partition a network. When you configured tap0 and
766 tap1 as access ports for different VLANs, you indicated that they
767 should be isolated from each other.
769 As for the external network and the Internet, it seems likely that
770 the machines you are trying to access are not on VLAN 9 (or 10) and
771 that the Internet is not available on VLAN 9 (or 10).
773 Q: Can I configure an IP address on a VLAN?
775 A: Yes. Use an "internal port" configured as an access port. For
776 example, the following configures IP address 192.168.0.7 on VLAN 9.
777 That is, OVS will forward packets from eth0 to 192.168.0.7 only if
778 they have an 802.1Q header with VLAN 9. Conversely, traffic
779 forwarded from 192.168.0.7 to eth0 will be tagged with an 802.1Q
783 ovs-vsctl add-port br0 eth0
784 ovs-vsctl add-port br0 vlan9 tag=9 -- set interface vlan9 type=internal
785 ifconfig vlan9 192.168.0.7
787 Q: My OpenFlow controller doesn't see the VLANs that I expect.
789 A: The configuration for VLANs in the Open vSwitch database (e.g. via
790 ovs-vsctl) only affects traffic that goes through Open vSwitch's
791 implementation of the OpenFlow "normal switching" action. By
792 default, when Open vSwitch isn't connected to a controller and
793 nothing has been manually configured in the flow table, all traffic
794 goes through the "normal switching" action. But, if you set up
795 OpenFlow flows on your own, through a controller or using ovs-ofctl
796 or through other means, then you have to implement VLAN handling
799 You can use "normal switching" as a component of your OpenFlow
800 actions, e.g. by putting "normal" into the lists of actions on
801 ovs-ofctl or by outputting to OFPP_NORMAL from an OpenFlow
802 controller. This will only be suitable for some situations,
805 Q: I configured ports on a bridge as access ports with different VLAN
809 ovs-vsctl set-controller br0 tcp:192.168.0.10:6633
810 ovs-vsctl add-port br0 eth0
811 ovs-vsctl add-port br0 tap0 tag=9
812 ovs-vsctl add-port br0 tap1 tag=10
814 but the VMs running behind tap0 and tap1 can still communicate,
815 that is, they are not isolated from each other even though they are
818 A: Do you have a controller configured on br0 (as the commands above
819 do)? If so, then this is a variant on the previous question, "My
820 OpenFlow controller doesn't see the VLANs that I expect," and you
821 can refer to the answer there for more information.
829 A: VXLAN stands for Virtual eXtensible Local Area Network, and is a means
830 to solve the scaling challenges of VLAN networks in a multi-tenant
831 environment. VXLAN is an overlay network which transports an L2 network
832 over an existing L3 network. For more information on VXLAN, please see
833 the IETF draft available here:
835 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03
837 Q: How much of the VXLAN protocol does Open vSwitch currently support?
839 A: Open vSwitch currently supports the framing format for packets on the
840 wire. There is currently no support for the multicast aspects of VXLAN.
841 To get around the lack of multicast support, it is possible to
842 pre-provision MAC to IP address mappings either manually or from a
845 Q: What destination UDP port does the VXLAN implementation in Open vSwitch
848 A: By default, Open vSwitch will use the assigned IANA port for VXLAN, which
849 is 4789. However, it is possible to configure the destination UDP port
850 manually on a per-VXLAN tunnel basis. An example of this configuration is
854 ovs-vsctl add-port br0 vxlan1 -- set interface vxlan1
855 type=vxlan options:remote_ip=192.168.1.2 options:key=flow
856 options:dst_port=8472
862 Q: What versions of OpenFlow does Open vSwitch support?
864 A: Open vSwitch 1.9 and earlier support only OpenFlow 1.0 (plus
865 extensions that bring in many of the features from later versions
868 Open vSwitch versions 1.10 and later will have experimental support
869 for OpenFlow 1.2 and 1.3. On these versions of Open vSwitch, the
870 following command enables OpenFlow 1.0, 1.2, and 1.3 on bridge br0:
872 ovs-vsctl set bridge br0 protocols=openflow10,openflow12,openflow13
874 Support for OpenFlow 1.1 is incomplete enough that it cannot yet be
875 enabled, even experimentally.
877 Support for OpenFlow 1.2 and 1.3 is still incomplete. Work to be
878 done is tracked in OPENFLOW-1.1+ in the Open vSwitch source tree
879 (also via http://openvswitch.org/development/openflow-1-x-plan/).
880 When support for a given OpenFlow version is solidly implemented,
881 Open vSwitch will enable that version by default.
883 Q: I'm getting "error type 45250 code 0". What's that?
885 A: This is a Open vSwitch extension to OpenFlow error codes. Open
886 vSwitch uses this extension when it must report an error to an
887 OpenFlow controller but no standard OpenFlow error code is
890 Open vSwitch logs the errors that it sends to controllers, so the
891 easiest thing to do is probably to look at the ovs-vswitchd log to
892 find out what the error was.
894 If you want to dissect the extended error message yourself, the
895 format is documented in include/openflow/nicira-ext.h in the Open
896 vSwitch source distribution. The extended error codes are
897 documented in lib/ofp-errors.h.
899 Q1: Some of the traffic that I'd expect my OpenFlow controller to see
900 doesn't actually appear through the OpenFlow connection, even
901 though I know that it's going through.
902 Q2: Some of the OpenFlow flows that my controller sets up don't seem
903 to apply to certain traffic, especially traffic between OVS and
904 the controller itself.
906 A: By default, Open vSwitch assumes that OpenFlow controllers are
907 connected "in-band", that is, that the controllers are actually
908 part of the network that is being controlled. In in-band mode,
909 Open vSwitch sets up special "hidden" flows to make sure that
910 traffic can make it back and forth between OVS and the controllers.
911 These hidden flows are higher priority than any flows that can be
912 set up through OpenFlow, and they are not visible through normal
913 OpenFlow flow table dumps.
915 Usually, the hidden flows are desirable and helpful, but
916 occasionally they can cause unexpected behavior. You can view the
917 full OpenFlow flow table, including hidden flows, on bridge br0
920 ovs-appctl bridge/dump-flows br0
922 to help you debug. The hidden flows are those with priorities
923 greater than 65535 (the maximum priority that can be set with
926 The DESIGN file at the top level of the Open vSwitch source
927 distribution describes the in-band model in detail.
929 If your controllers are not actually in-band (e.g. they are on
930 localhost via 127.0.0.1, or on a separate network), then you should
931 configure your controllers in "out-of-band" mode. If you have one
932 controller on bridge br0, then you can configure out-of-band mode
935 ovs-vsctl set controller br0 connection-mode=out-of-band
937 Q: I configured all my controllers for out-of-band control mode but
938 "ovs-appctl bridge/dump-flows" still shows some hidden flows.
940 A: You probably have a remote manager configured (e.g. with "ovs-vsctl
941 set-manager"). By default, Open vSwitch assumes that managers need
942 in-band rules set up on every bridge. You can disable these rules
945 ovs-vsctl set bridge br0 other-config:disable-in-band=true
947 This actually disables in-band control entirely for the bridge, as
948 if all the bridge's controllers were configured for out-of-band
951 Q: My OpenFlow controller doesn't see the VLANs that I expect.
953 A: See answer under "VLANs", above.
955 Q: I ran "ovs-ofctl add-flow br0 nw_dst=192.168.0.1,actions=drop"
956 but I got a funny message like this:
958 ofp_util|INFO|normalization changed ofp_match, details:
959 ofp_util|INFO| pre: nw_dst=192.168.0.1
962 and when I ran "ovs-ofctl dump-flows br0" I saw that my nw_dst
963 match had disappeared, so that the flow ends up matching every
966 A: The term "normalization" in the log message means that a flow
967 cannot match on an L3 field without saying what L3 protocol is in
968 use. The "ovs-ofctl" command above didn't specify an L3 protocol,
969 so the L3 field match was dropped.
971 In this case, the L3 protocol could be IP or ARP. A correct
972 command for each possibility is, respectively:
974 ovs-ofctl add-flow br0 ip,nw_dst=192.168.0.1,actions=drop
978 ovs-ofctl add-flow br0 arp,nw_dst=192.168.0.1,actions=drop
980 Similarly, a flow cannot match on an L4 field without saying what
981 L4 protocol is in use. For example, the flow match "tp_src=1234"
982 is, by itself, meaningless and will be ignored. Instead, to match
983 TCP source port 1234, write "tcp,tp_src=1234", or to match UDP
984 source port 1234, write "udp,tp_src=1234".
986 Q: How can I figure out the OpenFlow port number for a given port?
988 A: The OFPT_FEATURES_REQUEST message requests an OpenFlow switch to
989 respond with an OFPT_FEATURES_REPLY that, among other information,
990 includes a mapping between OpenFlow port names and numbers. From a
991 command prompt, "ovs-ofctl show br0" makes such a request and
992 prints the response for switch br0.
994 The Interface table in the Open vSwitch database also maps OpenFlow
995 port names to numbers. To print the OpenFlow port number
996 associated with interface eth0, run:
998 ovs-vsctl get Interface eth0 ofport
1000 You can print the entire mapping with:
1002 ovs-vsctl -- --columns=name,ofport list Interface
1004 but the output mixes together interfaces from all bridges in the
1005 database, so it may be confusing if more than one bridge exists.
1007 In the Open vSwitch database, ofport value -1 means that the
1008 interface could not be created due to an error. (The Open vSwitch
1009 log should indicate the reason.) ofport value [] (the empty set)
1010 means that the interface hasn't been created yet. The latter is
1011 normally an intermittent condition (unless ovs-vswitchd is not
1014 Q: I added some flows with my controller or with ovs-ofctl, but when I
1015 run "ovs-dpctl dump-flows" I don't see them.
1017 A: ovs-dpctl queries a kernel datapath, not an OpenFlow switch. It
1018 won't display the information that you want. You want to use
1019 "ovs-ofctl dump-flows" instead.
1021 Q: It looks like each of the interfaces in my bonded port shows up
1022 as an individual OpenFlow port. Is that right?
1024 A: Yes, Open vSwitch makes individual bond interfaces visible as
1025 OpenFlow ports, rather than the bond as a whole. The interfaces
1026 are treated together as a bond for only a few purposes:
1028 - Sending a packet to the OFPP_NORMAL port. (When an OpenFlow
1029 controller is not configured, this happens implicitly to
1032 - The "autopath" Nicira extension action. However, "autopath"
1033 is deprecated and scheduled for removal in February 2013.
1035 - Mirrors configured for output to a bonded port.
1037 It would make a lot of sense for Open vSwitch to present a bond as
1038 a single OpenFlow port. If you want to contribute an
1039 implementation of such a feature, please bring it up on the Open
1040 vSwitch development mailing list at dev@openvswitch.org.
1045 bugs@openvswitch.org
1046 http://openvswitch.org/
projects / cascardo / ovs.git / blob