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 2.6.18 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 ------------ -------------
153 Open vSwitch userspace should also work with the Linux kernel module
154 built into Linux 3.3 and later.
156 Open vSwitch userspace is not sensitive to the Linux kernel version.
157 It should build against almost any kernel, certainly against 2.6.18
160 Q: What Linux kernel versions does IPFIX flow monitoring work with?
162 A: IPFIX flow monitoring requires the Linux kernel module from Open
163 vSwitch version 1.10.90 or later.
165 Q: Should userspace or kernel be upgraded first to minimize downtime?
167 In general, the Open vSwitch userspace should be used with the
168 kernel version included in the same release or with the version
169 from upstream Linux. However, when upgrading between two releases
170 of Open vSwitch it is best to migrate userspace first to reduce
171 the possbility of incompatibilities.
173 Q: What features are not available in the Open vSwitch kernel datapath
174 that ships as part of the upstream Linux kernel?
176 A: The kernel module in upstream Linux 3.3 and later does not include
177 tunnel virtual ports, that is, interfaces with type "gre",
178 "ipsec_gre", "gre64", "ipsec_gre64", "vxlan", or "lisp". It is
179 possible to create tunnels in Linux and attach them to Open vSwitch
180 as system devices. However, they cannot be dynamically created
181 through the OVSDB protocol or set the tunnel ids as a flow action.
183 Work is in progress in adding tunnel virtual ports to the upstream
184 Linux version of the Open vSwitch kernel module. For now, if you
185 need these features, use the kernel module from the Open vSwitch
186 distribution instead of the upstream Linux kernel module.
188 The upstream kernel module does not include patch ports, but this
189 only matters for Open vSwitch 1.9 and earlier, because Open vSwitch
190 1.10 and later implement patch ports without using this kernel
193 Q: What features are not available when using the userspace datapath?
195 A: Tunnel virtual ports are not supported, as described in the
196 previous answer. It is also not possible to use queue-related
197 actions. On Linux kernels before 2.6.39, maximum-sized VLAN packets
198 may not be transmitted.
204 Q: I thought Open vSwitch was a virtual Ethernet switch, but the
205 documentation keeps talking about bridges. What's a bridge?
207 A: In networking, the terms "bridge" and "switch" are synonyms. Open
208 vSwitch implements an Ethernet switch, which means that it is also
213 A: See the "VLAN" section below.
219 Q: How do I configure a port as an access port?
221 A: Add "tag=VLAN" to your "ovs-vsctl add-port" command. For example,
222 the following commands configure br0 with eth0 as a trunk port (the
223 default) and tap0 as an access port for VLAN 9:
226 ovs-vsctl add-port br0 eth0
227 ovs-vsctl add-port br0 tap0 tag=9
229 If you want to configure an already added port as an access port,
230 use "ovs-vsctl set", e.g.:
232 ovs-vsctl set port tap0 tag=9
234 Q: How do I configure a port as a SPAN port, that is, enable mirroring
235 of all traffic to that port?
237 A: The following commands configure br0 with eth0 and tap0 as trunk
238 ports. All traffic coming in or going out on eth0 or tap0 is also
239 mirrored to tap1; any traffic arriving on tap1 is dropped:
242 ovs-vsctl add-port br0 eth0
243 ovs-vsctl add-port br0 tap0
244 ovs-vsctl add-port br0 tap1 \
245 -- --id=@p get port tap1 \
246 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
247 -- set bridge br0 mirrors=@m
249 To later disable mirroring, run:
251 ovs-vsctl clear bridge br0 mirrors
253 Q: How do I configure a VLAN as an RSPAN VLAN, that is, enable
254 mirroring of all traffic to that VLAN?
256 A: The following commands configure br0 with eth0 as a trunk port and
257 tap0 as an access port for VLAN 10. All traffic coming in or going
258 out on tap0, as well as traffic coming in or going out on eth0 in
259 VLAN 10, is also mirrored to VLAN 15 on eth0. The original tag for
260 VLAN 10, in cases where one is present, is dropped as part of
264 ovs-vsctl add-port br0 eth0
265 ovs-vsctl add-port br0 tap0 tag=10
267 -- --id=@m create mirror name=m0 select-all=true select-vlan=10 \
269 -- set bridge br0 mirrors=@m
271 To later disable mirroring, run:
273 ovs-vsctl clear bridge br0 mirrors
275 Mirroring to a VLAN can disrupt a network that contains unmanaged
276 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
277 GRE tunnel has fewer caveats than mirroring to a VLAN and should
278 generally be preferred.
280 Q: Can I mirror more than one input VLAN to an RSPAN VLAN?
282 A: Yes, but mirroring to a VLAN strips the original VLAN tag in favor
283 of the specified output-vlan. This loss of information may make
284 the mirrored traffic too hard to interpret.
286 To mirror multiple VLANs, use the commands above, but specify a
287 comma-separated list of VLANs as the value for select-vlan. To
288 mirror every VLAN, use the commands above, but omit select-vlan and
291 When a packet arrives on a VLAN that is used as a mirror output
292 VLAN, the mirror is disregarded. Instead, in standalone mode, OVS
293 floods the packet across all the ports for which the mirror output
294 VLAN is configured. (If an OpenFlow controller is in use, then it
295 can override this behavior through the flow table.) If OVS is used
296 as an intermediate switch, rather than an edge switch, this ensures
297 that the RSPAN traffic is distributed through the network.
299 Mirroring to a VLAN can disrupt a network that contains unmanaged
300 switches. See ovs-vswitchd.conf.db(5) for details. Mirroring to a
301 GRE tunnel has fewer caveats than mirroring to a VLAN and should
302 generally be preferred.
304 Q: How do I configure mirroring of all traffic to a GRE tunnel?
306 A: The following commands configure br0 with eth0 and tap0 as trunk
307 ports. All traffic coming in or going out on eth0 or tap0 is also
308 mirrored to gre0, a GRE tunnel to the remote host 192.168.1.10; any
309 traffic arriving on gre0 is dropped:
312 ovs-vsctl add-port br0 eth0
313 ovs-vsctl add-port br0 tap0
314 ovs-vsctl add-port br0 gre0 \
315 -- set interface gre0 type=gre options:remote_ip=192.168.1.10 \
316 -- --id=@p get port gre0 \
317 -- --id=@m create mirror name=m0 select-all=true output-port=@p \
318 -- set bridge br0 mirrors=@m
320 To later disable mirroring and destroy the GRE tunnel:
322 ovs-vsctl clear bridge br0 mirrors
323 ovs-vcstl del-port br0 gre0
325 Q: Does Open vSwitch support ERSPAN?
327 A: No. ERSPAN is an undocumented proprietary protocol. As an
328 alternative, Open vSwitch supports mirroring to a GRE tunnel (see
331 Q: How do I connect two bridges?
333 A: First, why do you want to do this? Two connected bridges are not
334 much different from a single bridge, so you might as well just have
335 a single bridge with all your ports on it.
337 If you still want to connect two bridges, you can use a pair of
338 patch ports. The following example creates bridges br0 and br1,
339 adds eth0 and tap0 to br0, adds tap1 to br1, and then connects br0
340 and br1 with a pair of patch ports.
343 ovs-vsctl add-port br0 eth0
344 ovs-vsctl add-port br0 tap0
346 ovs-vsctl add-port br1 tap1
348 -- add-port br0 patch0 \
349 -- set interface patch0 type=patch options:peer=patch1 \
350 -- add-port br1 patch1 \
351 -- set interface patch1 type=patch options:peer=patch0
353 Bridges connected with patch ports are much like a single bridge.
354 For instance, if the example above also added eth1 to br1, and both
355 eth0 and eth1 happened to be connected to the same next-hop switch,
356 then you could loop your network just as you would if you added
357 eth0 and eth1 to the same bridge (see the "Configuration Problems"
358 section below for more information).
360 If you are using Open vSwitch 1.9 or an earlier version, then you
361 need to be using the kernel module bundled with Open vSwitch rather
362 than the one that is integrated into Linux 3.3 and later, because
363 Open vSwitch 1.9 and earlier versions need kernel support for patch
364 ports. This also means that in Open vSwitch 1.9 and earlier, patch
365 ports will not work with the userspace datapath, only with the
368 Q: Why are there so many different ways to dump flows?
370 A: Open vSwitch uses different kinds of flows for different purposes:
372 - OpenFlow flows are the most important kind of flow. OpenFlow
373 controllers use these flows to define a switch's policy.
374 OpenFlow flows support wildcards, priorities, and multiple
377 When in-band control is in use, Open vSwitch sets up a few
378 "hidden" flows, with priority higher than a controller or the
379 user can configure, that are not visible via OpenFlow. (See
380 the "Controller" section of the FAQ for more information
383 - The Open vSwitch software switch implementation uses a second
384 kind of flow internally. These flows, called "exact-match"
385 or "datapath" or "kernel" flows, do not support wildcards or
386 priorities and comprise only a single table, which makes them
387 suitable for caching. OpenFlow flows and exact-match flows
388 also support different actions and number ports differently.
390 Exact-match flows are an implementation detail that is
391 subject to change in future versions of Open vSwitch. Even
392 with the current version of Open vSwitch, hardware switch
393 implementations do not necessarily use exact-match flows.
395 Each of the commands for dumping flows has a different purpose:
397 - "ovs-ofctl dump-flows <br>" dumps OpenFlow flows, excluding
398 hidden flows. This is the most commonly useful form of flow
399 dump. (Unlike the other commands, this should work with any
400 OpenFlow switch, not just Open vSwitch.)
402 - "ovs-appctl bridge/dump-flows <br>" dumps OpenFlow flows,
403 including hidden flows. This is occasionally useful for
404 troubleshooting suspected issues with in-band control.
406 - "ovs-dpctl dump-flows [dp]" dumps the exact-match flow table
407 entries for a Linux kernel-based datapath. In Open vSwitch
408 1.10 and later, ovs-vswitchd merges multiple switches into a
409 single datapath, so it will show all the flows on all your
410 kernel-based switches. This command can occasionally be
411 useful for debugging.
413 - "ovs-appctl dpif/dump-flows <br>", new in Open vSwitch 1.10,
414 dumps exact-match flows for only the specified bridge,
415 regardless of the type.
418 Configuration Problems
419 ----------------------
421 Q: I created a bridge and added my Ethernet port to it, using commands
425 ovs-vsctl add-port br0 eth0
427 and as soon as I ran the "add-port" command I lost all connectivity
430 A: A physical Ethernet device that is part of an Open vSwitch bridge
431 should not have an IP address. If one does, then that IP address
432 will not be fully functional.
434 You can restore functionality by moving the IP address to an Open
435 vSwitch "internal" device, such as the network device named after
436 the bridge itself. For example, assuming that eth0's IP address is
437 192.168.128.5, you could run the commands below to fix up the
440 ifconfig eth0 0.0.0.0
441 ifconfig br0 192.168.128.5
443 (If your only connection to the machine running OVS is through the
444 IP address in question, then you would want to run all of these
445 commands on a single command line, or put them into a script.) If
446 there were any additional routes assigned to eth0, then you would
447 also want to use commands to adjust these routes to go through br0.
449 If you use DHCP to obtain an IP address, then you should kill the
450 DHCP client that was listening on the physical Ethernet interface
451 (e.g. eth0) and start one listening on the internal interface
452 (e.g. br0). You might still need to manually clear the IP address
453 from the physical interface (e.g. with "ifconfig eth0 0.0.0.0").
455 There is no compelling reason why Open vSwitch must work this way.
456 However, this is the way that the Linux kernel bridge module has
457 always worked, so it's a model that those accustomed to Linux
458 bridging are already used to. Also, the model that most people
459 expect is not implementable without kernel changes on all the
460 versions of Linux that Open vSwitch supports.
462 By the way, this issue is not specific to physical Ethernet
463 devices. It applies to all network devices except Open vswitch
466 Q: I created a bridge and added a couple of Ethernet ports to it,
467 using commands like these:
470 ovs-vsctl add-port br0 eth0
471 ovs-vsctl add-port br0 eth1
473 and now my network seems to have melted: connectivity is unreliable
474 (even connectivity that doesn't go through Open vSwitch), all the
475 LEDs on my physical switches are blinking, wireshark shows
476 duplicated packets, and CPU usage is very high.
478 A: More than likely, you've looped your network. Probably, eth0 and
479 eth1 are connected to the same physical Ethernet switch. This
480 yields a scenario where OVS receives a broadcast packet on eth0 and
481 sends it out on eth1, then the physical switch connected to eth1
482 sends the packet back on eth0, and so on forever. More complicated
483 scenarios, involving a loop through multiple switches, are possible
486 The solution depends on what you are trying to do:
488 - If you added eth0 and eth1 to get higher bandwidth or higher
489 reliability between OVS and your physical Ethernet switch,
490 use a bond. The following commands create br0 and then add
491 eth0 and eth1 as a bond:
494 ovs-vsctl add-bond br0 bond0 eth0 eth1
496 Bonds have tons of configuration options. Please read the
497 documentation on the Port table in ovs-vswitchd.conf.db(5)
500 - Perhaps you don't actually need eth0 and eth1 to be on the
501 same bridge. For example, if you simply want to be able to
502 connect each of them to virtual machines, then you can put
503 each of them on a bridge of its own:
506 ovs-vsctl add-port br0 eth0
509 ovs-vsctl add-port br1 eth1
511 and then connect VMs to br0 and br1. (A potential
512 disadvantage is that traffic cannot directly pass between br0
513 and br1. Instead, it will go out eth0 and come back in eth1,
516 - If you have a redundant or complex network topology and you
517 want to prevent loops, turn on spanning tree protocol (STP).
518 The following commands create br0, enable STP, and add eth0
519 and eth1 to the bridge. The order is important because you
520 don't want have to have a loop in your network even
524 ovs-vsctl set bridge br0 stp_enable=true
525 ovs-vsctl add-port br0 eth0
526 ovs-vsctl add-port br0 eth1
528 The Open vSwitch implementation of STP is not well tested.
529 Please report any bugs you observe, but if you'd rather avoid
530 acting as a beta tester then another option might be your
533 Q: I can't seem to use Open vSwitch in a wireless network.
535 A: Wireless base stations generally only allow packets with the source
536 MAC address of NIC that completed the initial handshake.
537 Therefore, without MAC rewriting, only a single device can
538 communicate over a single wireless link.
540 This isn't specific to Open vSwitch, it's enforced by the access
541 point, so the same problems will show up with the Linux bridge or
542 any other way to do bridging.
544 Q: I can't seem to add my PPP interface to an Open vSwitch bridge.
546 A: PPP most commonly carries IP packets, but Open vSwitch works only
547 with Ethernet frames. The correct way to interface PPP to an
548 Ethernet network is usually to use routing instead of switching.
550 Q: Is there any documentation on the database tables and fields?
552 A: Yes. ovs-vswitchd.conf.db(5) is a comprehensive reference.
554 Q: When I run ovs-dpctl I no longer see the bridges I created. Instead,
555 I only see a datapath called "ovs-system". How can I see datapath
556 information about a particular bridge?
558 A: In version 1.9.0, OVS switched to using a single datapath that is
559 shared by all bridges of that type. The "ovs-appctl dpif/*"
560 commands provide similar functionality that is scoped by the bridge.
563 Quality of Service (QoS)
564 ------------------------
566 Q: How do I configure Quality of Service (QoS)?
568 A: Suppose that you want to set up bridge br0 connected to physical
569 Ethernet port eth0 (a 1 Gbps device) and virtual machine interfaces
570 vif1.0 and vif2.0, and that you want to limit traffic from vif1.0
571 to eth0 to 10 Mbps and from vif2.0 to eth0 to 20 Mbps. Then, you
572 could configure the bridge this way:
576 add-port br0 eth0 -- \
577 add-port br0 vif1.0 -- set interface vif1.0 ofport_request=5 -- \
578 add-port br0 vif2.0 -- set interface vif2.0 ofport_request=6 -- \
579 set port eth0 qos=@newqos -- \
580 --id=@newqos create qos type=linux-htb \
581 other-config:max-rate=1000000000 \
582 queues:123=@vif10queue \
583 queues:234=@vif20queue -- \
584 --id=@vif10queue create queue other-config:max-rate=10000000 -- \
585 --id=@vif20queue create queue other-config:max-rate=20000000
587 At this point, bridge br0 is configured with the ports and eth0 is
588 configured with the queues that you need for QoS, but nothing is
589 actually directing packets from vif1.0 or vif2.0 to the queues that
590 we have set up for them. That means that all of the packets to
591 eth0 are going to the "default queue", which is not what we want.
593 We use OpenFlow to direct packets from vif1.0 and vif2.0 to the
594 queues reserved for them:
596 ovs-ofctl add-flow br0 in_port=5,actions=set_queue:123,normal
597 ovs-ofctl add-flow br0 in_port=6,actions=set_queue:234,normal
599 Each of the above flows matches on the input port, sets up the
600 appropriate queue (123 for vif1.0, 234 for vif2.0), and then
601 executes the "normal" action, which performs the same switching
602 that Open vSwitch would have done without any OpenFlow flows being
603 present. (We know that vif1.0 and vif2.0 have OpenFlow port
604 numbers 5 and 6, respectively, because we set their ofport_request
605 columns above. If we had not done that, then we would have needed
606 to find out their port numbers before setting up these flows.)
608 Now traffic going from vif1.0 or vif2.0 to eth0 should be
611 By the way, if you delete the bridge created by the above commands,
616 then that will leave one unreferenced QoS record and two
617 unreferenced Queue records in the Open vSwich database. One way to
618 clear them out, assuming you don't have other QoS or Queue records
619 that you want to keep, is:
621 ovs-vsctl -- --all destroy QoS -- --all destroy Queue
623 If you do want to keep some QoS or Queue records, or the Open
624 vSwitch you are using is older than version 1.8 (which added the
625 --all option), then you will have to destroy QoS and Queue records
628 Q: I configured Quality of Service (QoS) in my OpenFlow network by
629 adding records to the QoS and Queue table, but the results aren't
632 A: Did you install OpenFlow flows that use your queues? This is the
633 primary way to tell Open vSwitch which queues you want to use. If
634 you don't do this, then the default queue will be used, which will
635 probably not have the effect you want.
637 Refer to the previous question for an example.
639 Q: I configured QoS, correctly, but my measurements show that it isn't
640 working as well as I expect.
642 A: With the Linux kernel, the Open vSwitch implementation of QoS has
645 - Open vSwitch configures a subset of Linux kernel QoS
646 features, according to what is in OVSDB. It is possible that
647 this code has bugs. If you believe that this is so, then you
648 can configure the Linux traffic control (QoS) stack directly
649 with the "tc" program. If you get better results that way,
650 you can send a detailed bug report to bugs@openvswitch.org.
652 It is certain that Open vSwitch cannot configure every Linux
653 kernel QoS feature. If you need some feature that OVS cannot
654 configure, then you can also use "tc" directly (or add that
657 - The Open vSwitch implementation of OpenFlow allows flows to
658 be directed to particular queues. This is pretty simple and
659 unlikely to have serious bugs at this point.
661 However, most problems with QoS on Linux are not bugs in Open
662 vSwitch at all. They tend to be either configuration errors
663 (please see the earlier questions in this section) or issues with
664 the traffic control (QoS) stack in Linux. The Open vSwitch
665 developers are not experts on Linux traffic control. We suggest
666 that, if you believe you are encountering a problem with Linux
667 traffic control, that you consult the tc manpages (e.g. tc(8),
668 tc-htb(8), tc-hfsc(8)), web resources (e.g. http://lartc.org/), or
669 mailing lists (e.g. http://vger.kernel.org/vger-lists.html#netdev).
677 A: At the simplest level, a VLAN (short for "virtual LAN") is a way to
678 partition a single switch into multiple switches. Suppose, for
679 example, that you have two groups of machines, group A and group B.
680 You want the machines in group A to be able to talk to each other,
681 and you want the machine in group B to be able to talk to each
682 other, but you don't want the machines in group A to be able to
683 talk to the machines in group B. You can do this with two
684 switches, by plugging the machines in group A into one switch and
685 the machines in group B into the other switch.
687 If you only have one switch, then you can use VLANs to do the same
688 thing, by configuring the ports for machines in group A as VLAN
689 "access ports" for one VLAN and the ports for group B as "access
690 ports" for a different VLAN. The switch will only forward packets
691 between ports that are assigned to the same VLAN, so this
692 effectively subdivides your single switch into two independent
693 switches, one for each group of machines.
695 So far we haven't said anything about VLAN headers. With access
696 ports, like we've described so far, no VLAN header is present in
697 the Ethernet frame. This means that the machines (or switches)
698 connected to access ports need not be aware that VLANs are
699 involved, just like in the case where we use two different physical
702 Now suppose that you have a whole bunch of switches in your
703 network, instead of just one, and that some machines in group A are
704 connected directly to both switches 1 and 2. To allow these
705 machines to talk to each other, you could add an access port for
706 group A's VLAN to switch 1 and another to switch 2, and then
707 connect an Ethernet cable between those ports. That works fine,
708 but it doesn't scale well as the number of switches and the number
709 of VLANs increases, because you use up a lot of valuable switch
710 ports just connecting together your VLANs.
712 This is where VLAN headers come in. Instead of using one cable and
713 two ports per VLAN to connect a pair of switches, we configure a
714 port on each switch as a VLAN "trunk port". Packets sent and
715 received on a trunk port carry a VLAN header that says what VLAN
716 the packet belongs to, so that only two ports total are required to
717 connect the switches, regardless of the number of VLANs in use.
718 Normally, only switches (either physical or virtual) are connected
719 to a trunk port, not individual hosts, because individual hosts
720 don't expect to see a VLAN header in the traffic that they receive.
722 None of the above discussion says anything about particular VLAN
723 numbers. This is because VLAN numbers are completely arbitrary.
724 One must only ensure that a given VLAN is numbered consistently
725 throughout a network and that different VLANs are given different
726 numbers. (That said, VLAN 0 is usually synonymous with a packet
727 that has no VLAN header, and VLAN 4095 is reserved.)
731 A: Many drivers in Linux kernels before version 3.3 had VLAN-related
732 bugs. If you are having problems with VLANs that you suspect to be
733 driver related, then you have several options:
735 - Upgrade to Linux 3.3 or later.
737 - Build and install a fixed version of the particular driver
738 that is causing trouble, if one is available.
740 - Use a NIC whose driver does not have VLAN problems.
742 - Use "VLAN splinters", a feature in Open vSwitch 1.4 and later
743 that works around bugs in kernel drivers. To enable VLAN
744 splinters on interface eth0, use the command:
746 ovs-vsctl set interface eth0 other-config:enable-vlan-splinters=true
748 For VLAN splinters to be effective, Open vSwitch must know
749 which VLANs are in use. See the "VLAN splinters" section in
750 the Interface table in ovs-vswitchd.conf.db(5) for details on
751 how Open vSwitch infers in-use VLANs.
753 VLAN splinters increase memory use and reduce performance, so
754 use them only if needed.
756 - Apply the "vlan workaround" patch from the XenServer kernel
757 patch queue, build Open vSwitch against this patched kernel,
758 and then use ovs-vlan-bug-workaround(8) to enable the VLAN
759 workaround for each interface whose driver is buggy.
761 (This is a nontrivial exercise, so this option is included
762 only for completeness.)
764 It is not always easy to tell whether a Linux kernel driver has
765 buggy VLAN support. The ovs-vlan-test(8) and ovs-test(8) utilities
766 can help you test. See their manpages for details. Of the two
767 utilities, ovs-test(8) is newer and more thorough, but
768 ovs-vlan-test(8) may be easier to use.
770 Q: VLANs still don't work. I've tested the driver so I know that it's OK.
772 A: Do you have VLANs enabled on the physical switch that OVS is
773 attached to? Make sure that the port is configured to trunk the
774 VLAN or VLANs that you are using with OVS.
776 Q: Outgoing VLAN-tagged traffic goes through OVS to my physical switch
777 and to its destination host, but OVS seems to drop incoming return
780 A: It's possible that you have the VLAN configured on your physical
781 switch as the "native" VLAN. In this mode, the switch treats
782 incoming packets either tagged with the native VLAN or untagged as
783 part of the native VLAN. It may also send outgoing packets in the
784 native VLAN without a VLAN tag.
786 If this is the case, you have two choices:
788 - Change the physical switch port configuration to tag packets
789 it forwards to OVS with the native VLAN instead of forwarding
792 - Change the OVS configuration for the physical port to a
793 native VLAN mode. For example, the following sets up a
794 bridge with port eth0 in "native-tagged" mode in VLAN 9:
797 ovs-vsctl add-port br0 eth0 tag=9 vlan_mode=native-tagged
799 In this situation, "native-untagged" mode will probably work
800 equally well. Refer to the documentation for the Port table
801 in ovs-vswitchd.conf.db(5) for more information.
803 Q: I added a pair of VMs on different VLANs, like this:
806 ovs-vsctl add-port br0 eth0
807 ovs-vsctl add-port br0 tap0 tag=9
808 ovs-vsctl add-port br0 tap1 tag=10
810 but the VMs can't access each other, the external network, or the
813 A: It is to be expected that the VMs can't access each other. VLANs
814 are a means to partition a network. When you configured tap0 and
815 tap1 as access ports for different VLANs, you indicated that they
816 should be isolated from each other.
818 As for the external network and the Internet, it seems likely that
819 the machines you are trying to access are not on VLAN 9 (or 10) and
820 that the Internet is not available on VLAN 9 (or 10).
822 Q: I added a pair of VMs on the same VLAN, like this:
825 ovs-vsctl add-port br0 eth0
826 ovs-vsctl add-port br0 tap0 tag=9
827 ovs-vsctl add-port br0 tap1 tag=9
829 The VMs can access each other, but not the external network or the
832 A: It seems likely that the machines you are trying to access in the
833 external network are not on VLAN 9 and that the Internet is not
834 available on VLAN 9. Also, ensure VLAN 9 is set up as an allowed
835 trunk VLAN on the upstream switch port to which eth0 is connected.
837 Q: Can I configure an IP address on a VLAN?
839 A: Yes. Use an "internal port" configured as an access port. For
840 example, the following configures IP address 192.168.0.7 on VLAN 9.
841 That is, OVS will forward packets from eth0 to 192.168.0.7 only if
842 they have an 802.1Q header with VLAN 9. Conversely, traffic
843 forwarded from 192.168.0.7 to eth0 will be tagged with an 802.1Q
847 ovs-vsctl add-port br0 eth0
848 ovs-vsctl add-port br0 vlan9 tag=9 -- set interface vlan9 type=internal
849 ifconfig vlan9 192.168.0.7
851 Q: My OpenFlow controller doesn't see the VLANs that I expect.
853 A: The configuration for VLANs in the Open vSwitch database (e.g. via
854 ovs-vsctl) only affects traffic that goes through Open vSwitch's
855 implementation of the OpenFlow "normal switching" action. By
856 default, when Open vSwitch isn't connected to a controller and
857 nothing has been manually configured in the flow table, all traffic
858 goes through the "normal switching" action. But, if you set up
859 OpenFlow flows on your own, through a controller or using ovs-ofctl
860 or through other means, then you have to implement VLAN handling
863 You can use "normal switching" as a component of your OpenFlow
864 actions, e.g. by putting "normal" into the lists of actions on
865 ovs-ofctl or by outputting to OFPP_NORMAL from an OpenFlow
866 controller. In situations where this is not suitable, you can
867 implement VLAN handling yourself, e.g.:
869 - If a packet comes in on an access port, and the flow table
870 needs to send it out on a trunk port, then the flow can add
871 the appropriate VLAN tag with the "mod_vlan_vid" action.
873 - If a packet comes in on a trunk port, and the flow table
874 needs to send it out on an access port, then the flow can
875 strip the VLAN tag with the "strip_vlan" action.
877 Q: I configured ports on a bridge as access ports with different VLAN
881 ovs-vsctl set-controller br0 tcp:192.168.0.10:6633
882 ovs-vsctl add-port br0 eth0
883 ovs-vsctl add-port br0 tap0 tag=9
884 ovs-vsctl add-port br0 tap1 tag=10
886 but the VMs running behind tap0 and tap1 can still communicate,
887 that is, they are not isolated from each other even though they are
890 A: Do you have a controller configured on br0 (as the commands above
891 do)? If so, then this is a variant on the previous question, "My
892 OpenFlow controller doesn't see the VLANs that I expect," and you
893 can refer to the answer there for more information.
901 A: VXLAN stands for Virtual eXtensible Local Area Network, and is a means
902 to solve the scaling challenges of VLAN networks in a multi-tenant
903 environment. VXLAN is an overlay network which transports an L2 network
904 over an existing L3 network. For more information on VXLAN, please see
905 the IETF draft available here:
907 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03
909 Q: How much of the VXLAN protocol does Open vSwitch currently support?
911 A: Open vSwitch currently supports the framing format for packets on the
912 wire. There is currently no support for the multicast aspects of VXLAN.
913 To get around the lack of multicast support, it is possible to
914 pre-provision MAC to IP address mappings either manually or from a
917 Q: What destination UDP port does the VXLAN implementation in Open vSwitch
920 A: By default, Open vSwitch will use the assigned IANA port for VXLAN, which
921 is 4789. However, it is possible to configure the destination UDP port
922 manually on a per-VXLAN tunnel basis. An example of this configuration is
926 ovs-vsctl add-port br0 vxlan1 -- set interface vxlan1
927 type=vxlan options:remote_ip=192.168.1.2 options:key=flow
928 options:dst_port=8472
931 Using OpenFlow (Manually or Via Controller)
932 -------------------------------------------
934 Q: What versions of OpenFlow does Open vSwitch support?
936 A: Open vSwitch 1.9 and earlier support only OpenFlow 1.0 (plus
937 extensions that bring in many of the features from later versions
940 Open vSwitch 1.10 and later have experimental support for OpenFlow
941 1.2 and 1.3. On these versions of Open vSwitch, the following
942 command enables OpenFlow 1.0, 1.2, and 1.3 on bridge br0:
944 ovs-vsctl set bridge br0 protocols=OpenFlow10,OpenFlow12,OpenFlow13
946 Open vSwitch version 1.12 and later will have experimental support
947 for OpenFlow 1.1, 1.2, and 1.3. On these versions of Open vSwitch,
948 the following command enables OpenFlow 1.0, 1.1, 1.2, and 1.3 on
951 ovs-vsctl set bridge br0 protocols=OpenFlow10,OpenFlow11,OpenFlow12,OpenFlow13
953 Use the -O option to enable support for later versions of OpenFlow
954 in ovs-ofctl. For example:
956 ovs-ofctl -O OpenFlow13 dump-flows br0
958 Support for OpenFlow 1.1, 1.2, and 1.3 is still incomplete. Work
959 to be done is tracked in OPENFLOW-1.1+ in the Open vSwitch sources
960 (also via http://openvswitch.org/development/openflow-1-x-plan/).
961 When support for a given OpenFlow version is solidly implemented,
962 Open vSwitch will enable that version by default.
964 Q: I'm getting "error type 45250 code 0". What's that?
966 A: This is a Open vSwitch extension to OpenFlow error codes. Open
967 vSwitch uses this extension when it must report an error to an
968 OpenFlow controller but no standard OpenFlow error code is
971 Open vSwitch logs the errors that it sends to controllers, so the
972 easiest thing to do is probably to look at the ovs-vswitchd log to
973 find out what the error was.
975 If you want to dissect the extended error message yourself, the
976 format is documented in include/openflow/nicira-ext.h in the Open
977 vSwitch source distribution. The extended error codes are
978 documented in lib/ofp-errors.h.
980 Q1: Some of the traffic that I'd expect my OpenFlow controller to see
981 doesn't actually appear through the OpenFlow connection, even
982 though I know that it's going through.
983 Q2: Some of the OpenFlow flows that my controller sets up don't seem
984 to apply to certain traffic, especially traffic between OVS and
985 the controller itself.
987 A: By default, Open vSwitch assumes that OpenFlow controllers are
988 connected "in-band", that is, that the controllers are actually
989 part of the network that is being controlled. In in-band mode,
990 Open vSwitch sets up special "hidden" flows to make sure that
991 traffic can make it back and forth between OVS and the controllers.
992 These hidden flows are higher priority than any flows that can be
993 set up through OpenFlow, and they are not visible through normal
994 OpenFlow flow table dumps.
996 Usually, the hidden flows are desirable and helpful, but
997 occasionally they can cause unexpected behavior. You can view the
998 full OpenFlow flow table, including hidden flows, on bridge br0
1001 ovs-appctl bridge/dump-flows br0
1003 to help you debug. The hidden flows are those with priorities
1004 greater than 65535 (the maximum priority that can be set with
1007 The DESIGN file at the top level of the Open vSwitch source
1008 distribution describes the in-band model in detail.
1010 If your controllers are not actually in-band (e.g. they are on
1011 localhost via 127.0.0.1, or on a separate network), then you should
1012 configure your controllers in "out-of-band" mode. If you have one
1013 controller on bridge br0, then you can configure out-of-band mode
1016 ovs-vsctl set controller br0 connection-mode=out-of-band
1018 Q: I configured all my controllers for out-of-band control mode but
1019 "ovs-appctl bridge/dump-flows" still shows some hidden flows.
1021 A: You probably have a remote manager configured (e.g. with "ovs-vsctl
1022 set-manager"). By default, Open vSwitch assumes that managers need
1023 in-band rules set up on every bridge. You can disable these rules
1026 ovs-vsctl set bridge br0 other-config:disable-in-band=true
1028 This actually disables in-band control entirely for the bridge, as
1029 if all the bridge's controllers were configured for out-of-band
1032 Q: My OpenFlow controller doesn't see the VLANs that I expect.
1034 A: See answer under "VLANs", above.
1036 Q: I ran "ovs-ofctl add-flow br0 nw_dst=192.168.0.1,actions=drop"
1037 but I got a funny message like this:
1039 ofp_util|INFO|normalization changed ofp_match, details:
1040 ofp_util|INFO| pre: nw_dst=192.168.0.1
1043 and when I ran "ovs-ofctl dump-flows br0" I saw that my nw_dst
1044 match had disappeared, so that the flow ends up matching every
1047 A: The term "normalization" in the log message means that a flow
1048 cannot match on an L3 field without saying what L3 protocol is in
1049 use. The "ovs-ofctl" command above didn't specify an L3 protocol,
1050 so the L3 field match was dropped.
1052 In this case, the L3 protocol could be IP or ARP. A correct
1053 command for each possibility is, respectively:
1055 ovs-ofctl add-flow br0 ip,nw_dst=192.168.0.1,actions=drop
1059 ovs-ofctl add-flow br0 arp,nw_dst=192.168.0.1,actions=drop
1061 Similarly, a flow cannot match on an L4 field without saying what
1062 L4 protocol is in use. For example, the flow match "tp_src=1234"
1063 is, by itself, meaningless and will be ignored. Instead, to match
1064 TCP source port 1234, write "tcp,tp_src=1234", or to match UDP
1065 source port 1234, write "udp,tp_src=1234".
1067 Q: How can I figure out the OpenFlow port number for a given port?
1069 A: The OFPT_FEATURES_REQUEST message requests an OpenFlow switch to
1070 respond with an OFPT_FEATURES_REPLY that, among other information,
1071 includes a mapping between OpenFlow port names and numbers. From a
1072 command prompt, "ovs-ofctl show br0" makes such a request and
1073 prints the response for switch br0.
1075 The Interface table in the Open vSwitch database also maps OpenFlow
1076 port names to numbers. To print the OpenFlow port number
1077 associated with interface eth0, run:
1079 ovs-vsctl get Interface eth0 ofport
1081 You can print the entire mapping with:
1083 ovs-vsctl -- --columns=name,ofport list Interface
1085 but the output mixes together interfaces from all bridges in the
1086 database, so it may be confusing if more than one bridge exists.
1088 In the Open vSwitch database, ofport value -1 means that the
1089 interface could not be created due to an error. (The Open vSwitch
1090 log should indicate the reason.) ofport value [] (the empty set)
1091 means that the interface hasn't been created yet. The latter is
1092 normally an intermittent condition (unless ovs-vswitchd is not
1095 Q: I added some flows with my controller or with ovs-ofctl, but when I
1096 run "ovs-dpctl dump-flows" I don't see them.
1098 A: ovs-dpctl queries a kernel datapath, not an OpenFlow switch. It
1099 won't display the information that you want. You want to use
1100 "ovs-ofctl dump-flows" instead.
1102 Q: It looks like each of the interfaces in my bonded port shows up
1103 as an individual OpenFlow port. Is that right?
1105 A: Yes, Open vSwitch makes individual bond interfaces visible as
1106 OpenFlow ports, rather than the bond as a whole. The interfaces
1107 are treated together as a bond for only a few purposes:
1109 - Sending a packet to the OFPP_NORMAL port. (When an OpenFlow
1110 controller is not configured, this happens implicitly to
1113 - Mirrors configured for output to a bonded port.
1115 It would make a lot of sense for Open vSwitch to present a bond as
1116 a single OpenFlow port. If you want to contribute an
1117 implementation of such a feature, please bring it up on the Open
1118 vSwitch development mailing list at dev@openvswitch.org.
1120 Q: I have a sophisticated network setup involving Open vSwitch, VMs or
1121 multiple hosts, and other components. The behavior isn't what I
1124 A: To debug network behavior problems, trace the path of a packet,
1125 hop-by-hop, from its origin in one host to a remote host. If
1126 that's correct, then trace the path of the response packet back to
1129 Usually a simple ICMP echo request and reply ("ping") packet is
1130 good enough. Start by initiating an ongoing "ping" from the origin
1131 host to a remote host. If you are tracking down a connectivity
1132 problem, the "ping" will not display any successful output, but
1133 packets are still being sent. (In this case the packets being sent
1134 are likely ARP rather than ICMP.)
1136 Tools available for tracing include the following:
1138 - "tcpdump" and "wireshark" for observing hops across network
1139 devices, such as Open vSwitch internal devices and physical
1142 - "ovs-appctl dpif/dump-flows <br>" in Open vSwitch 1.10 and
1143 later or "ovs-dpctl dump-flows <br>" in earlier versions.
1144 These tools allow one to observe the actions being taken on
1145 packets in ongoing flows.
1147 See ovs-vswitchd(8) for "ovs-appctl dpif/dump-flows"
1148 documentation, ovs-dpctl(8) for "ovs-dpctl dump-flows"
1149 documentation, and "Why are there so many different ways to
1150 dump flows?" above for some background.
1152 - "ovs-appctl ofproto/trace" to observe the logic behind how
1153 ovs-vswitchd treats packets. See ovs-vswitchd(8) for
1154 documentation. You can out more details about a given flow
1155 that "ovs-dpctl dump-flows" displays, by cutting and pasting
1156 a flow from the output into an "ovs-appctl ofproto/trace"
1159 - SPAN, RSPAN, and ERSPAN features of physical switches, to
1160 observe what goes on at these physical hops.
1162 Starting at the origin of a given packet, observe the packet at
1163 each hop in turn. For example, in one plausible scenario, you
1166 1. "tcpdump" the "eth" interface through which an ARP egresses
1167 a VM, from inside the VM.
1169 2. "tcpdump" the "vif" or "tap" interface through which the ARP
1170 ingresses the host machine.
1172 3. Use "ovs-dpctl dump-flows" to spot the ARP flow and observe
1173 the host interface through which the ARP egresses the
1174 physical machine. You may need to use "ovs-dpctl show" to
1175 interpret the port numbers. If the output seems surprising,
1176 you can use "ovs-appctl ofproto/trace" to observe details of
1177 how ovs-vswitchd determined the actions in the "ovs-dpctl
1180 4. "tcpdump" the "eth" interface through which the ARP egresses
1181 the physical machine.
1183 5. "tcpdump" the "eth" interface through which the ARP
1184 ingresses the physical machine, at the remote host that
1187 6. Use "ovs-dpctl dump-flows" to spot the ARP flow on the
1188 remote host that receives the ARP and observe the VM "vif"
1189 or "tap" interface to which the flow is directed. Again,
1190 "ovs-dpctl show" and "ovs-appctl ofproto/trace" might help.
1192 7. "tcpdump" the "vif" or "tap" interface to which the ARP is
1195 8. "tcpdump" the "eth" interface through which the ARP
1196 ingresses a VM, from inside the VM.
1198 It is likely that during one of these steps you will figure out the
1199 problem. If not, then follow the ARP reply back to the origin, in
1202 Q: How do I make a flow drop packets?
1204 A: An empty set of actions causes a packet to be dropped. You can
1205 specify an empty set of actions with "actions=" on the ovs-ofctl
1206 command line. For example:
1208 ovs-ofctl add-flow br0 priority=65535,actions=
1210 would cause every packet entering switch br0 to be dropped.
1212 You can write "drop" explicitly if you like. The effect is the
1213 same. Thus, the following command also causes every packet
1214 entering switch br0 to be dropped:
1216 ovs-ofctl add-flow br0 priority=65535,actions=drop
1222 bugs@openvswitch.org
1223 http://openvswitch.org/
git://git.onelab.eu / sliver-openvswitch.git / blob