1 <?xml version="1.0" encoding="utf-8"?>
2 <database title="Open vSwitch Configuration Database">
4 A database with this schema holds the configuration for one Open
5 vSwitch daemon. The top-level configuration for the daemon is the
6 <ref table="Open_vSwitch"/> table, which must have exactly one
7 record. Records in other tables are significant only when they
8 can be reached directly or indirectly from the <ref
9 table="Open_vSwitch"/> table. Records that are not reachable from
10 the <ref table="Open_vSwitch"/> table are automatically deleted
11 from the database, except for records in a few distinguished
15 <h2>Common Columns</h2>
18 Most tables contain two special columns, named <code>other_config</code>
19 and <code>external_ids</code>. These columns have the same form and
20 purpose each place that they appear, so we describe them here to save space
25 <dt><code>other_config</code>: map of string-string pairs</dt>
28 Key-value pairs for configuring rarely used features. Supported keys,
29 along with the forms taken by their values, are documented individually
33 A few tables do not have <code>other_config</code> columns because no
34 key-value pairs have yet been defined for them.
38 <dt><code>external_ids</code>: map of string-string pairs</dt>
40 Key-value pairs for use by external frameworks that integrate with Open
41 vSwitch, rather than by Open vSwitch itself. System integrators should
42 either use the Open vSwitch development mailing list to coordinate on
43 common key-value definitions, or choose key names that are likely to be
44 unique. In some cases, where key-value pairs have been defined that are
45 likely to be widely useful, they are documented individually for each
50 <table name="Open_vSwitch" title="Open vSwitch configuration.">
51 Configuration for an Open vSwitch daemon. There must be exactly
52 one record in the <ref table="Open_vSwitch"/> table.
54 <group title="Configuration">
55 <column name="bridges">
56 Set of bridges managed by the daemon.
60 SSL used globally by the daemon.
63 <column name="external_ids" key="system-id">
64 A unique identifier for the Open vSwitch's physical host.
65 The form of the identifier depends on the type of the host.
66 On a Citrix XenServer, this will likely be the same as
67 <ref column="external_ids" key="xs-system-uuid"/>.
70 <column name="external_ids" key="xs-system-uuid">
71 The Citrix XenServer universally unique identifier for the physical
72 host as displayed by <code>xe host-list</code>.
76 <group title="Status">
77 <column name="next_cfg">
78 Sequence number for client to increment. When a client modifies
79 any part of the database configuration and wishes to wait for
80 Open vSwitch to finish applying the changes, it may increment
84 <column name="cur_cfg">
85 Sequence number that Open vSwitch sets to the current value of
86 <ref column="next_cfg"/> after it finishes applying a set of
87 configuration changes.
90 <column name="capabilities">
91 Describes functionality supported by the hardware and software platform
92 on which this Open vSwitch is based. Clients should not modify this
93 column. See the <ref table="Capability"/> description for defined
94 capability categories and the meaning of associated
95 <ref table="Capability"/> records.
98 <group title="Statistics">
100 The <code>statistics</code> column contains key-value pairs that
101 report statistics about a system running an Open vSwitch. These are
102 updated periodically (currently, every 5 seconds). Key-value pairs
103 that cannot be determined or that do not apply to a platform are
107 <column name="other_config" key="enable-statistics"
108 type='{"type": "boolean"}'>
109 Statistics are disabled by default to avoid overhead in the common
110 case when statistics gathering is not useful. Set this value to
111 <code>true</code> to enable populating the <ref column="statistics"/>
112 column or to <code>false</code> to explicitly disable it.
115 <column name="statistics" key="cpu"
116 type='{"type": "integer", "minInteger": 1}'>
118 Number of CPU processors, threads, or cores currently online and
119 available to the operating system on which Open vSwitch is running,
120 as an integer. This may be less than the number installed, if some
121 are not online or if they are not available to the operating
125 Open vSwitch userspace processes are not multithreaded, but the
126 Linux kernel-based datapath is.
130 <column name="statistics" key="load_average">
131 A comma-separated list of three floating-point numbers,
132 representing the system load average over the last 1, 5, and 15
133 minutes, respectively.
136 <column name="statistics" key="memory">
138 A comma-separated list of integers, each of which represents a
139 quantity of memory in kilobytes that describes the operating
140 system on which Open vSwitch is running. In respective order,
145 <li>Total amount of RAM allocated to the OS.</li>
146 <li>RAM allocated to the OS that is in use.</li>
147 <li>RAM that can be flushed out to disk or otherwise discarded
148 if that space is needed for another purpose. This number is
149 necessarily less than or equal to the previous value.</li>
150 <li>Total disk space allocated for swap.</li>
151 <li>Swap space currently in use.</li>
155 On Linux, all five values can be determined and are included. On
156 other operating systems, only the first two values can be
157 determined, so the list will only have two values.
161 <column name="statistics" key="process_NAME">
163 One such key-value pair, with <code>NAME</code> replaced by
164 a process name, will exist for each running Open vSwitch
165 daemon process, with <var>name</var> replaced by the
166 daemon's name (e.g. <code>process_ovs-vswitchd</code>). The
167 value is a comma-separated list of integers. The integers
168 represent the following, with memory measured in kilobytes
169 and durations in milliseconds:
173 <li>The process's virtual memory size.</li>
174 <li>The process's resident set size.</li>
175 <li>The amount of user and system CPU time consumed by the
177 <li>The number of times that the process has crashed and been
178 automatically restarted by the monitor.</li>
179 <li>The duration since the process was started.</li>
180 <li>The duration for which the process has been running.</li>
184 The interpretation of some of these values depends on whether the
185 process was started with the <option>--monitor</option>. If it
186 was not, then the crash count will always be 0 and the two
187 durations will always be the same. If <option>--monitor</option>
188 was given, then the crash count may be positive; if it is, the
189 latter duration is the amount of time since the most recent crash
194 There will be one key-value pair for each file in Open vSwitch's
195 ``run directory'' (usually <code>/var/run/openvswitch</code>)
196 whose name ends in <code>.pid</code>, whose contents are a
197 process ID, and which is locked by a running process. The
198 <var>name</var> is taken from the pidfile's name.
202 Currently Open vSwitch is only able to obtain all of the above
203 detail on Linux systems. On other systems, the same key-value
204 pairs will be present but the values will always be the empty
209 <column name="statistics" key="file_systems">
211 A space-separated list of information on local, writable file
212 systems. Each item in the list describes one file system and
213 consists in turn of a comma-separated list of the following:
217 <li>Mount point, e.g. <code>/</code> or <code>/var/log</code>.
218 Any spaces or commas in the mount point are replaced by
220 <li>Total size, in kilobytes, as an integer.</li>
221 <li>Amount of storage in use, in kilobytes, as an integer.</li>
225 This key-value pair is omitted if there are no local, writable
226 file systems or if Open vSwitch cannot obtain the needed
233 <group title="Version Reporting">
235 These columns report the types and versions of the hardware and
236 software running Open vSwitch. We recommend in general that software
237 should test whether specific features are supported instead of relying
238 on version number checks. These values are primarily intended for
239 reporting to human administrators.
242 <column name="ovs_version">
243 The Open vSwitch version number, e.g. <code>1.1.0</code>.
244 If Open vSwitch was configured with a build number, then it is
245 also included, e.g. <code>1.1.0+build6579</code>.
248 <column name="db_version">
250 The database schema version number in the form
251 <code><var>major</var>.<var>minor</var>.<var>tweak</var></code>,
252 e.g. <code>1.2.3</code>. Whenever the database schema is changed in
253 a non-backward compatible way (e.g. deleting a column or a table),
254 <var>major</var> is incremented. When the database schema is changed
255 in a backward compatible way (e.g. adding a new column),
256 <var>minor</var> is incremented. When the database schema is changed
257 cosmetically (e.g. reindenting its syntax), <var>tweak</var> is
262 The schema version is part of the database schema, so it can also be
263 retrieved by fetching the schema using the Open vSwitch database
268 <column name="system_type">
270 An identifier for the type of system on top of which Open vSwitch
271 runs, e.g. <code>XenServer</code> or <code>KVM</code>.
274 System integrators are responsible for choosing and setting an
275 appropriate value for this column.
279 <column name="system_version">
281 The version of the system identified by <ref column="system_type"/>,
282 e.g. <code>5.6.100-39265p</code> on XenServer 5.6.100 build 39265.
285 System integrators are responsible for choosing and setting an
286 appropriate value for this column.
292 <group title="Database Configuration">
294 These columns primarily configure the Open vSwitch database
295 (<code>ovsdb-server</code>), not the Open vSwitch switch
296 (<code>ovs-vswitchd</code>). The OVSDB database also uses the <ref
297 column="ssl"/> settings.
301 The Open vSwitch switch does read the database configuration to
302 determine remote IP addresses to which in-band control should apply.
305 <column name="manager_options">
306 Database clients to which the Open vSwitch database server should
307 connect or to which it should listen, along with options for how these
308 connection should be configured. See the <ref table="Manager"/> table
309 for more information.
313 <group title="Common Columns">
314 The overall purpose of these columns is described under <code>Common
315 Columns</code> at the beginning of this document.
317 <column name="other_config"/>
318 <column name="external_ids"/>
322 <table name="Bridge">
324 Configuration for a bridge within an
325 <ref table="Open_vSwitch"/>.
328 A <ref table="Bridge"/> record represents an Ethernet switch with one or
329 more ``ports,'' which are the <ref table="Port"/> records pointed to by
330 the <ref table="Bridge"/>'s <ref column="ports"/> column.
333 <group title="Core Features">
335 Bridge identifier. Should be alphanumeric and no more than about 8
336 bytes long. Must be unique among the names of ports, interfaces, and
340 <column name="ports">
341 Ports included in the bridge.
344 <column name="mirrors">
345 Port mirroring configuration.
348 <column name="netflow">
349 NetFlow configuration.
352 <column name="sflow">
356 <column name="flood_vlans">
358 VLAN IDs of VLANs on which MAC address learning should be disabled,
359 so that packets are flooded instead of being sent to specific ports
360 that are believed to contain packets' destination MACs. This should
361 ordinarily be used to disable MAC learning on VLANs used for
362 mirroring (RSPAN VLANs). It may also be useful for debugging.
365 SLB bonding (see the <ref table="Port" column="bond_mode"/> column in
366 the <ref table="Port"/> table) is incompatible with
367 <code>flood_vlans</code>. Consider using another bonding mode or
368 a different type of mirror instead.
373 <group title="OpenFlow Configuration">
374 <column name="controller">
376 OpenFlow controller set. If unset, then no OpenFlow controllers
381 If there are primary controllers, removing all of them clears the
382 flow table. If there are no primary controllers, adding one also
383 clears the flow table. Other changes to the set of controllers, such
384 as adding or removing a service controller, adding another primary
385 controller to supplement an existing primary controller, or removing
386 only one of two primary controllers, have no effect on the flow
391 <column name="flow_tables">
392 Configuration for OpenFlow tables. Each pair maps from an OpenFlow
393 table ID to configuration for that table.
396 <column name="fail_mode">
397 <p>When a controller is configured, it is, ordinarily, responsible
398 for setting up all flows on the switch. Thus, if the connection to
399 the controller fails, no new network connections can be set up.
400 If the connection to the controller stays down long enough,
401 no packets can pass through the switch at all. This setting
402 determines the switch's response to such a situation. It may be set
403 to one of the following:
405 <dt><code>standalone</code></dt>
406 <dd>If no message is received from the controller for three
407 times the inactivity probe interval
408 (see <ref column="inactivity_probe"/>), then Open vSwitch
409 will take over responsibility for setting up flows. In
410 this mode, Open vSwitch causes the bridge to act like an
411 ordinary MAC-learning switch. Open vSwitch will continue
412 to retry connecting to the controller in the background
413 and, when the connection succeeds, it will discontinue its
414 standalone behavior.</dd>
415 <dt><code>secure</code></dt>
416 <dd>Open vSwitch will not set up flows on its own when the
417 controller connection fails or when no controllers are
418 defined. The bridge will continue to retry connecting to
419 any defined controllers forever.</dd>
422 <p>If this value is unset, the default is implementation-specific.</p>
423 <p>When more than one controller is configured,
424 <ref column="fail_mode"/> is considered only when none of the
425 configured controllers can be contacted.</p>
427 Changing <ref column="fail_mode"/> when no primary controllers are
428 configured clears the flow table.
432 <column name="datapath_id">
433 Reports the OpenFlow datapath ID in use. Exactly 16 hex digits.
434 (Setting this column has no useful effect. Set <ref
435 column="other-config" key="datapath-id"/> instead.)
438 <column name="other_config" key="datapath-id">
439 Exactly 16 hex digits to set the OpenFlow datapath ID to a specific
440 value. May not be all-zero.
443 <column name="other_config" key="disable-in-band"
444 type='{"type": "boolean"}'>
445 If set to <code>true</code>, disable in-band control on the bridge
446 regardless of controller and manager settings.
449 <column name="other_config" key="in-band-queue"
450 type='{"type": "integer", "minInteger": 0, "maxInteger": 4294967295}'>
451 A queue ID as a nonnegative integer. This sets the OpenFlow queue ID
452 that will be used by flows set up by in-band control on this bridge.
453 If unset, or if the port used by an in-band control flow does not have
454 QoS configured, or if the port does not have a queue with the specified
455 ID, the default queue is used instead.
459 <group title="Spanning Tree Configuration">
460 The IEEE 802.1D Spanning Tree Protocol (STP) is a network protocol
461 that ensures loop-free topologies. It allows redundant links to
462 be included in the network to provide automatic backup paths if
463 the active links fails.
465 <column name="stp_enable">
466 Enable spanning tree on the bridge. By default, STP is disabled
467 on bridges. Bond, internal, and mirror ports are not supported
468 and will not participate in the spanning tree.
471 <column name="other_config" key="stp-system-id">
472 The bridge's STP identifier (the lower 48 bits of the bridge-id)
474 <var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>.
475 By default, the identifier is the MAC address of the bridge.
478 <column name="other_config" key="stp-priority"
479 type='{"type": "integer", "minInteger": 0, "maxInteger": 65535}'>
480 The bridge's relative priority value for determining the root
481 bridge (the upper 16 bits of the bridge-id). A bridge with the
482 lowest bridge-id is elected the root. By default, the priority
486 <column name="other_config" key="stp-hello-time"
487 type='{"type": "integer", "minInteger": 1, "maxInteger": 10}'>
488 The interval between transmissions of hello messages by
489 designated ports, in seconds. By default the hello interval is
493 <column name="other_config" key="stp-max-age"
494 type='{"type": "integer", "minInteger": 6, "maxInteger": 40}'>
495 The maximum age of the information transmitted by the bridge
496 when it is the root bridge, in seconds. By default, the maximum
500 <column name="other_config" key="stp-forward-delay"
501 type='{"type": "integer", "minInteger": 4, "maxInteger": 30}'>
502 The delay to wait between transitioning root and designated
503 ports to <code>forwarding</code>, in seconds. By default, the
504 forwarding delay is 15 seconds.
508 <group title="Other Features">
509 <column name="datapath_type">
510 Name of datapath provider. The kernel datapath has
511 type <code>system</code>. The userspace datapath has
512 type <code>netdev</code>.
515 <column name="external_ids" key="bridge-id">
516 A unique identifier of the bridge. On Citrix XenServer this will
517 commonly be the same as
518 <ref column="external_ids" key="xs-network-uuids"/>.
521 <column name="external_ids" key="xs-network-uuids">
522 Semicolon-delimited set of universally unique identifier(s) for the
523 network with which this bridge is associated on a Citrix XenServer
524 host. The network identifiers are RFC 4122 UUIDs as displayed by,
525 e.g., <code>xe network-list</code>.
528 <column name="other_config" key="hwaddr">
529 An Ethernet address in the form
530 <var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>
531 to set the hardware address of the local port and influence the
535 <column name="other_config" key="flow-eviction-threshold"
536 type='{"type": "integer", "minInteger": 0}'>
538 A number of flows as a nonnegative integer. This sets number of
539 flows at which eviction from the kernel flow table will be triggered.
540 If there are a large number of flows then increasing this value to
541 around the number of flows present can result in reduced CPU usage
545 The default is 1000. Values below 100 will be rounded up to 100.
549 <column name="other_config" key="forward-bpdu"
550 type='{"type": "boolean"}'>
551 Option to allow forwarding of BPDU frames when NORMAL action is
552 invoked. Frames with reserved Ethernet addresses (e.g. STP
553 BPDU) will be forwarded when this option is enabled and the
554 switch is not providing that functionality. If STP is enabled
555 on the port, STP BPDUs will never be forwarded. If the Open
556 vSwitch bridge is used to connect different Ethernet networks,
557 and if Open vSwitch node does not run STP, then this option
558 should be enabled. Default is disabled, set to
559 <code>true</code> to enable.
562 <column name="other_config" key="mac-aging-time"
563 type='{"type": "integer", "minInteger": 1}'>
565 The maximum number of seconds to retain a MAC learning entry for
566 which no packets have been seen. The default is currently 300
567 seconds (5 minutes). The value, if specified, is forced into a
568 reasonable range, currently 15 to 3600 seconds.
572 A short MAC aging time allows a network to more quickly detect that a
573 host is no longer connected to a switch port. However, it also makes
574 it more likely that packets will be flooded unnecessarily, when they
575 are addressed to a connected host that rarely transmits packets. To
576 reduce the incidence of unnecessary flooding, use a MAC aging time
577 longer than the maximum interval at which a host will ordinarily
583 <group title="Bridge Status">
585 Status information about bridges.
587 <column name="status">
588 Key-value pairs that report bridge status.
590 <column name="status" key="stp_bridge_id">
592 The bridge-id (in hex) used in spanning tree advertisements.
593 Configuring the bridge-id is described in the
594 <code>stp-system-id</code> and <code>stp-priority</code> keys
595 of the <code>other_config</code> section earlier.
598 <column name="status" key="stp_designated_root">
600 The designated root (in hex) for this spanning tree.
603 <column name="status" key="stp_root_path_cost">
605 The path cost of reaching the designated bridge. A lower
611 <group title="Common Columns">
612 The overall purpose of these columns is described under <code>Common
613 Columns</code> at the beginning of this document.
615 <column name="other_config"/>
616 <column name="external_ids"/>
620 <table name="Port" table="Port or bond configuration.">
621 <p>A port within a <ref table="Bridge"/>.</p>
622 <p>Most commonly, a port has exactly one ``interface,'' pointed to by its
623 <ref column="interfaces"/> column. Such a port logically
624 corresponds to a port on a physical Ethernet switch. A port
625 with more than one interface is a ``bonded port'' (see
626 <ref group="Bonding Configuration"/>).</p>
627 <p>Some properties that one might think as belonging to a port are actually
628 part of the port's <ref table="Interface"/> members.</p>
631 Port name. Should be alphanumeric and no more than about 8
632 bytes long. May be the same as the interface name, for
633 non-bonded ports. Must otherwise be unique among the names of
634 ports, interfaces, and bridges on a host.
637 <column name="interfaces">
638 The port's interfaces. If there is more than one, this is a
642 <group title="VLAN Configuration">
643 <p>Bridge ports support the following types of VLAN configuration:</p>
648 A trunk port carries packets on one or more specified VLANs
649 specified in the <ref column="trunks"/> column (often, on every
650 VLAN). A packet that ingresses on a trunk port is in the VLAN
651 specified in its 802.1Q header, or VLAN 0 if the packet has no
652 802.1Q header. A packet that egresses through a trunk port will
653 have an 802.1Q header if it has a nonzero VLAN ID.
657 Any packet that ingresses on a trunk port tagged with a VLAN that
658 the port does not trunk is dropped.
665 An access port carries packets on exactly one VLAN specified in the
666 <ref column="tag"/> column. Packets egressing on an access port
667 have no 802.1Q header.
671 Any packet with an 802.1Q header with a nonzero VLAN ID that
672 ingresses on an access port is dropped, regardless of whether the
673 VLAN ID in the header is the access port's VLAN ID.
677 <dt>native-tagged</dt>
679 A native-tagged port resembles a trunk port, with the exception that
680 a packet without an 802.1Q header that ingresses on a native-tagged
681 port is in the ``native VLAN'' (specified in the <ref column="tag"/>
685 <dt>native-untagged</dt>
687 A native-untagged port resembles a native-tagged port, with the
688 exception that a packet that egresses on a native-untagged port in
689 the native VLAN will not have an 802.1Q header.
693 A packet will only egress through bridge ports that carry the VLAN of
694 the packet, as described by the rules above.
697 <column name="vlan_mode">
699 The VLAN mode of the port, as described above. When this column is
700 empty, a default mode is selected as follows:
704 If <ref column="tag"/> contains a value, the port is an access
705 port. The <ref column="trunks"/> column should be empty.
708 Otherwise, the port is a trunk port. The <ref column="trunks"/>
709 column value is honored if it is present.
716 For an access port, the port's implicitly tagged VLAN. For a
717 native-tagged or native-untagged port, the port's native VLAN. Must
718 be empty if this is a trunk port.
722 <column name="trunks">
724 For a trunk, native-tagged, or native-untagged port, the 802.1Q VLAN
725 or VLANs that this port trunks; if it is empty, then the port trunks
726 all VLANs. Must be empty if this is an access port.
729 A native-tagged or native-untagged port always trunks its native
730 VLAN, regardless of whether <ref column="trunks"/> includes that
735 <column name="other_config" key="priority-tags"
736 type='{"type": "boolean"}'>
738 An 802.1Q header contains two important pieces of information: a VLAN
739 ID and a priority. A frame with a zero VLAN ID, called a
740 ``priority-tagged'' frame, is supposed to be treated the same way as
741 a frame without an 802.1Q header at all (except for the priority).
745 However, some network elements ignore any frame that has 802.1Q
746 header at all, even when the VLAN ID is zero. Therefore, by default
747 Open vSwitch does not output priority-tagged frames, instead omitting
748 the 802.1Q header entirely if the VLAN ID is zero. Set this key to
749 <code>true</code> to enable priority-tagged frames on a port.
753 Regardless of this setting, Open vSwitch omits the 802.1Q header on
754 output if both the VLAN ID and priority would be zero.
758 All frames output to native-tagged ports have a nonzero VLAN ID, so
759 this setting is not meaningful on native-tagged ports.
764 <group title="Bonding Configuration">
765 <p>A port that has more than one interface is a ``bonded port.'' Bonding
766 allows for load balancing and fail-over. Some kinds of bonding will
767 work with any kind of upstream switch:</p>
770 <dt><code>balance-slb</code></dt>
772 Balances flows among slaves based on source MAC address and output
773 VLAN, with periodic rebalancing as traffic patterns change.
776 <dt><code>active-backup</code></dt>
778 Assigns all flows to one slave, failing over to a backup slave when
779 the active slave is disabled.
784 The following modes require the upstream switch to support 802.3ad with
785 successful LACP negotiation:
789 <dt><code>balance-tcp</code></dt>
791 Balances flows among slaves based on L2, L3, and L4 protocol
792 information such as destination MAC address, IP address, and TCP
796 <dt><code>stable</code></dt>
798 <p>Attempts to always assign a given flow to the same slave
799 consistently. In an effort to maintain stability, no load
800 balancing is done. Uses a similar hashing strategy to
801 <code>balance-tcp</code>, always taking into account L3 and L4
802 fields even if LACP negotiations are unsuccessful. </p>
803 <p>Slave selection decisions are made based on <ref table="Interface"
804 column="other_config" key="bond-stable-id"/> if set. Otherwise,
805 OpenFlow port number is used. Decisions are consistent across all
806 <code>ovs-vswitchd</code> instances with equivalent
807 <ref table="Interface" column="other_config" key="bond-stable-id"/>
812 <p>These columns apply only to bonded ports. Their values are
813 otherwise ignored.</p>
815 <column name="bond_mode">
816 <p>The type of bonding used for a bonded port. Defaults to
817 <code>active-backup</code> if unset.
821 <column name="other_config" key="bond-hash-basis"
822 type='{"type": "integer"}'>
823 An integer hashed along with flows when choosing output slaves in load
824 balanced bonds. When changed, all flows will be assigned different
825 hash values possibly causing slave selection decisions to change. Does
826 not affect bonding modes which do not employ load balancing such as
827 <code>active-backup</code>.
830 <group title="Link Failure Detection">
832 An important part of link bonding is detecting that links are down so
833 that they may be disabled. These settings determine how Open vSwitch
834 detects link failure.
837 <column name="other_config" key="bond-detect-mode"
838 type='{"type": "string", "enum": ["set", ["carrier", "miimon"]]}'>
839 The means used to detect link failures. Defaults to
840 <code>carrier</code> which uses each interface's carrier to detect
841 failures. When set to <code>miimon</code>, will check for failures
842 by polling each interface's MII.
845 <column name="other_config" key="bond-miimon-interval"
846 type='{"type": "integer"}'>
847 The interval, in milliseconds, between successive attempts to poll
848 each interface's MII. Relevant only when <ref column="other_config"
849 key="bond-detect-mode"/> is <code>miimon</code>.
852 <column name="bond_updelay">
854 The number of milliseconds for which carrier must stay up on an
855 interface before the interface is considered to be up. Specify
856 <code>0</code> to enable the interface immediately.
860 This setting is honored only when at least one bonded interface is
861 already enabled. When no interfaces are enabled, then the first
862 bond interface to come up is enabled immediately.
866 <column name="bond_downdelay">
867 The number of milliseconds for which carrier must stay down on an
868 interface before the interface is considered to be down. Specify
869 <code>0</code> to disable the interface immediately.
873 <group title="LACP Configuration">
875 LACP, the Link Aggregation Control Protocol, is an IEEE standard that
876 allows switches to automatically detect that they are connected by
877 multiple links and aggregate across those links. These settings
878 control LACP behavior.
882 Configures LACP on this port. LACP allows directly connected
883 switches to negotiate which links may be bonded. LACP may be enabled
884 on non-bonded ports for the benefit of any switches they may be
885 connected to. <code>active</code> ports are allowed to initiate LACP
886 negotiations. <code>passive</code> ports are allowed to participate
887 in LACP negotiations initiated by a remote switch, but not allowed to
888 initiate such negotiations themselves. If LACP is enabled on a port
889 whose partner switch does not support LACP, the bond will be
890 disabled. Defaults to <code>off</code> if unset.
893 <column name="other_config" key="lacp-system-id">
894 The LACP system ID of this <ref table="Port"/>. The system ID of a
895 LACP bond is used to identify itself to its partners. Must be a
896 nonzero MAC address. Defaults to the bridge Ethernet address if
900 <column name="other_config" key="lacp-system-priority"
901 type='{"type": "integer", "minInteger": 1, "maxInteger": 65535}'>
902 The LACP system priority of this <ref table="Port"/>. In LACP
903 negotiations, link status decisions are made by the system with the
904 numerically lower priority.
907 <column name="other_config" key="lacp-time">
909 The LACP timing which should be used on this <ref table="Port"/>.
910 Possible values are <code>fast</code>, <code>slow</code> and a
911 positive number of milliseconds. By default <code>slow</code> is
912 used. When configured to be <code>fast</code> LACP heartbeats are
913 requested at a rate of once per second causing connectivity
914 problems to be detected more quickly. In <code>slow</code> mode,
915 heartbeats are requested at a rate of once every 30 seconds.
919 Users may manually set a heartbeat transmission rate to increase
920 the fault detection speed further. When manually set, OVS expects
921 the partner switch to be configured with the same transmission
922 rate. Manually setting <code>lacp-time</code> to something other
923 than <code>fast</code> or <code>slow</code> is not supported by the
928 <column name="other_config" key="lacp-heartbeat"
929 type='{"type": "boolean"}'>
930 Treat LACP like a simple heartbeat protocol for link state
931 monitoring. Most features of the LACP protocol are disabled
932 when this mode is in use. The default if not specified is
937 <group title="SLB Configuration">
939 These settings control behavior when a bond is in
940 <code>balance-slb</code> mode, regardless of whether the bond was
941 intentionally configured in SLB mode or it fell back to SLB mode
942 because LACP negotiation failed.
945 <column name="other_config" key="bond-rebalance-interval"
946 type='{"type": "integer", "minInteger": 0, "maxInteger": 10000}'>
947 For a load balanced bonded port, the number of milliseconds between
948 successive attempts to rebalance the bond, that is, to move flows
949 from one interface on the bond to another in an attempt to keep usage
950 of each interface roughly equal. If zero, load balancing is disabled
951 on the bond (carrier status changes still cause flows to move). If
952 less than 1000ms, the rebalance interval will be 1000ms.
956 <column name="bond_fake_iface">
957 For a bonded port, whether to create a fake internal interface with the
958 name of the port. Use only for compatibility with legacy software that
963 <group title="Spanning Tree Configuration">
964 <column name="other_config" key="stp-enable"
965 type='{"type": "boolean"}'>
966 If spanning tree is enabled on the bridge, member ports are
967 enabled by default (with the exception of bond, internal, and
968 mirror ports which do not work with STP). If this column's
969 value is <code>false</code> spanning tree is disabled on the
973 <column name="other_config" key="stp-port-num"
974 type='{"type": "integer", "minInteger": 1, "maxInteger": 255}'>
975 The port number used for the lower 8 bits of the port-id. By
976 default, the numbers will be assigned automatically. If any
977 port's number is manually configured on a bridge, then they
981 <column name="other_config" key="stp-port-priority"
982 type='{"type": "integer", "minInteger": 0, "maxInteger": 255}'>
983 The port's relative priority value for determining the root
984 port (the upper 8 bits of the port-id). A port with a lower
985 port-id will be chosen as the root port. By default, the
989 <column name="other_config" key="stp-path-cost"
990 type='{"type": "integer", "minInteger": 0, "maxInteger": 65535}'>
991 Spanning tree path cost for the port. A lower number indicates
992 a faster link. By default, the cost is based on the maximum
997 <group title="Other Features">
999 Quality of Service configuration for this port.
1003 The MAC address to use for this port for the purpose of choosing the
1004 bridge's MAC address. This column does not necessarily reflect the
1005 port's actual MAC address, nor will setting it change the port's actual
1009 <column name="fake_bridge">
1010 Does this port represent a sub-bridge for its tagged VLAN within the
1011 Bridge? See ovs-vsctl(8) for more information.
1014 <column name="external_ids" key="fake-bridge-id-*">
1015 External IDs for a fake bridge (see the <ref column="fake_bridge"/>
1016 column) are defined by prefixing a <ref table="Bridge"/> <ref
1017 table="Bridge" column="external_ids"/> key with
1018 <code>fake-bridge-</code>,
1019 e.g. <code>fake-bridge-xs-network-uuids</code>.
1023 <group title="Port Status">
1025 Status information about ports attached to bridges.
1027 <column name="status">
1028 Key-value pairs that report port status.
1030 <column name="status" key="stp_port_id">
1032 The port-id (in hex) used in spanning tree advertisements for
1033 this port. Configuring the port-id is described in the
1034 <code>stp-port-num</code> and <code>stp-port-priority</code>
1035 keys of the <code>other_config</code> section earlier.
1038 <column name="status" key="stp_state"
1039 type='{"type": "string", "enum": ["set",
1040 ["disabled", "listening", "learning",
1041 "forwarding", "blocking"]]}'>
1043 STP state of the port.
1046 <column name="status" key="stp_sec_in_state"
1047 type='{"type": "integer", "minInteger": 0}'>
1049 The amount of time (in seconds) port has been in the current
1053 <column name="status" key="stp_role"
1054 type='{"type": "string", "enum": ["set",
1055 ["root", "designated", "alternate"]]}'>
1057 STP role of the port.
1062 <group title="Port Statistics">
1064 Key-value pairs that report port statistics.
1066 <group title="Statistics: STP transmit and receive counters">
1067 <column name="statistics" key="stp_tx_count">
1068 Number of STP BPDUs sent on this port by the spanning
1071 <column name="statistics" key="stp_rx_count">
1072 Number of STP BPDUs received on this port and accepted by the
1073 spanning tree library.
1075 <column name="statistics" key="stp_error_count">
1076 Number of bad STP BPDUs received on this port. Bad BPDUs
1077 include runt packets and those with an unexpected protocol ID.
1082 <group title="Common Columns">
1083 The overall purpose of these columns is described under <code>Common
1084 Columns</code> at the beginning of this document.
1086 <column name="other_config"/>
1087 <column name="external_ids"/>
1091 <table name="Interface" title="One physical network device in a Port.">
1092 An interface within a <ref table="Port"/>.
1094 <group title="Core Features">
1095 <column name="name">
1096 Interface name. Should be alphanumeric and no more than about 8 bytes
1097 long. May be the same as the port name, for non-bonded ports. Must
1098 otherwise be unique among the names of ports, interfaces, and bridges
1103 <p>Ethernet address to set for this interface. If unset then the
1104 default MAC address is used:</p>
1106 <li>For the local interface, the default is the lowest-numbered MAC
1107 address among the other bridge ports, either the value of the
1108 <ref table="Port" column="mac"/> in its <ref table="Port"/> record,
1109 if set, or its actual MAC (for bonded ports, the MAC of its slave
1110 whose name is first in alphabetical order). Internal ports and
1111 bridge ports that are used as port mirroring destinations (see the
1112 <ref table="Mirror"/> table) are ignored.</li>
1113 <li>For other internal interfaces, the default MAC is randomly
1115 <li>External interfaces typically have a MAC address associated with
1116 their hardware.</li>
1118 <p>Some interfaces may not have a software-controllable MAC
1122 <column name="ofport">
1123 <p>OpenFlow port number for this interface. Unlike most columns, this
1124 column's value should be set only by Open vSwitch itself. Other
1125 clients should set this column to an empty set (the default) when
1126 creating an <ref table="Interface"/>.</p>
1127 <p>Open vSwitch populates this column when the port number becomes
1128 known. If the interface is successfully added,
1129 <ref column="ofport"/> will be set to a number between 1 and 65535
1130 (generally either in the range 1 to 65279, inclusive, or 65534, the
1131 port number for the OpenFlow ``local port''). If the interface
1132 cannot be added then Open vSwitch sets this column
1137 <group title="System-Specific Details">
1138 <column name="type">
1140 The interface type, one of:
1144 <dt><code>system</code></dt>
1145 <dd>An ordinary network device, e.g. <code>eth0</code> on Linux.
1146 Sometimes referred to as ``external interfaces'' since they are
1147 generally connected to hardware external to that on which the Open
1148 vSwitch is running. The empty string is a synonym for
1149 <code>system</code>.</dd>
1151 <dt><code>internal</code></dt>
1152 <dd>A simulated network device that sends and receives traffic. An
1153 internal interface whose <ref column="name"/> is the same as its
1154 bridge's <ref table="Open_vSwitch" column="name"/> is called the
1155 ``local interface.'' It does not make sense to bond an internal
1156 interface, so the terms ``port'' and ``interface'' are often used
1157 imprecisely for internal interfaces.</dd>
1159 <dt><code>tap</code></dt>
1160 <dd>A TUN/TAP device managed by Open vSwitch.</dd>
1162 <dt><code>gre</code></dt>
1164 An Ethernet over RFC 2890 Generic Routing Encapsulation over IPv4
1165 tunnel. See <ref group="Tunnel Options"/> for information on
1166 configuring GRE tunnels.
1169 <dt><code>ipsec_gre</code></dt>
1171 An Ethernet over RFC 2890 Generic Routing Encapsulation over IPv4
1175 <dt><code>capwap</code></dt>
1177 An Ethernet tunnel over the UDP transport portion of CAPWAP (RFC
1178 5415). This allows interoperability with certain switches that do
1179 not support GRE. Only the tunneling component of the protocol is
1180 implemented. UDP ports 58881 and 58882 are used as the source and
1181 destination ports respectively. CAPWAP is currently supported only
1182 with the Linux kernel datapath with kernel version 2.6.26 or later.
1185 <dt><code>patch</code></dt>
1187 A pair of virtual devices that act as a patch cable.
1190 <dt><code>null</code></dt>
1191 <dd>An ignored interface.</dd>
1196 <group title="Tunnel Options">
1198 These options apply to interfaces with <ref column="type"/> of
1199 <code>gre</code>, <code>ipsec_gre</code>, and <code>capwap</code>.
1203 Each tunnel must be uniquely identified by the combination of <ref
1204 column="type"/>, <ref column="options" key="remote_ip"/>, <ref
1205 column="options" key="local_ip"/>, and <ref column="options"
1206 key="in_key"/>. If two ports are defined that are the same except one
1207 has an optional identifier and the other does not, the more specific
1208 one is matched first. <ref column="options" key="in_key"/> is
1209 considered more specific than <ref column="options" key="local_ip"/> if
1210 a port defines one and another port defines the other.
1213 <column name="options" key="remote_ip">
1215 Required. The tunnel endpoint. Unicast and multicast endpoints are
1220 When a multicast endpoint is specified, a routing table lookup occurs
1221 only when the tunnel is created. Following a routing change, delete
1222 and then re-create the tunnel to force a new routing table lookup.
1226 <column name="options" key="local_ip">
1227 Optional. The destination IP that received packets must match.
1228 Default is to match all addresses. Must be omitted when <ref
1229 column="options" key="remote_ip"/> is a multicast address.
1232 <column name="options" key="in_key">
1233 <p>Optional. The key that received packets must contain, one of:</p>
1237 <code>0</code>. The tunnel receives packets with no key or with a
1238 key of 0. This is equivalent to specifying no <ref column="options"
1239 key="in_key"/> at all.
1242 A positive 32-bit (for GRE) or 64-bit (for CAPWAP) number. The
1243 tunnel receives only packets with the specified key.
1246 The word <code>flow</code>. The tunnel accepts packets with any
1247 key. The key will be placed in the <code>tun_id</code> field for
1248 matching in the flow table. The <code>ovs-ofctl</code> manual page
1249 contains additional information about matching fields in OpenFlow
1258 <column name="options" key="out_key">
1259 <p>Optional. The key to be set on outgoing packets, one of:</p>
1263 <code>0</code>. Packets sent through the tunnel will have no key.
1264 This is equivalent to specifying no <ref column="options"
1265 key="out_key"/> at all.
1268 A positive 32-bit (for GRE) or 64-bit (for CAPWAP) number. Packets
1269 sent through the tunnel will have the specified key.
1272 The word <code>flow</code>. Packets sent through the tunnel will
1273 have the key set using the <code>set_tunnel</code> Nicira OpenFlow
1274 vendor extension (0 is used in the absence of an action). The
1275 <code>ovs-ofctl</code> manual page contains additional information
1276 about the Nicira OpenFlow vendor extensions.
1281 <column name="options" key="key">
1282 Optional. Shorthand to set <code>in_key</code> and
1283 <code>out_key</code> at the same time.
1286 <column name="options" key="tos">
1287 Optional. The value of the ToS bits to be set on the encapsulating
1288 packet. It may also be the word <code>inherit</code>, in which case
1289 the ToS will be copied from the inner packet if it is IPv4 or IPv6
1290 (otherwise it will be 0). The ECN fields are always inherited.
1294 <column name="options" key="ttl">
1295 Optional. The TTL to be set on the encapsulating packet. It may also
1296 be the word <code>inherit</code>, in which case the TTL will be copied
1297 from the inner packet if it is IPv4 or IPv6 (otherwise it will be the
1298 system default, typically 64). Default is the system default TTL.
1301 <column name="options" key="df_inherit" type='{"type": "boolean"}'>
1302 Optional. If enabled, the Don't Fragment bit will be copied from the
1303 inner IP headers (those of the encapsulated traffic) to the outer
1304 (tunnel) headers. Default is disabled; set to <code>true</code> to
1308 <column name="options" key="df_default"
1309 type='{"type": "boolean"}'>
1310 Optional. If enabled, the Don't Fragment bit will be set by default on
1311 tunnel headers if the <code>df_inherit</code> option is not set, or if
1312 the encapsulated packet is not IP. Default is enabled; set to
1313 <code>false</code> to disable.
1316 <column name="options" key="pmtud" type='{"type": "boolean"}'>
1317 Optional. Enable tunnel path MTU discovery. If enabled ``ICMP
1318 Destination Unreachable - Fragmentation Needed'' messages will be
1319 generated for IPv4 packets with the DF bit set and IPv6 packets above
1320 the minimum MTU if the packet size exceeds the path MTU minus the size
1321 of the tunnel headers. Note that this option causes behavior that is
1322 typically reserved for routers and therefore is not entirely in
1323 compliance with the IEEE 802.1D specification for bridges. Default is
1324 enabled; set to <code>false</code> to disable.
1327 <group title="Tunnel Options: gre only">
1329 Only <code>gre</code> interfaces support these options.
1332 <column name="options" key="header_cache" type='{"type": "boolean"}'>
1333 Enable caching of tunnel headers and the output path. This can lead
1334 to a significant performance increase without changing behavior. In
1335 general it should not be necessary to adjust this setting. However,
1336 the caching can bypass certain components of the IP stack (such as
1337 <code>iptables</code>) and it may be useful to disable it if these
1338 features are required or as a debugging measure. Default is enabled,
1339 set to <code>false</code> to disable.
1343 <group title="Tunnel Options: gre and ipsec_gre only">
1345 Only <code>gre</code> and <code>ipsec_gre</code> interfaces support
1349 <column name="options" key="csum" type='{"type": "boolean"}'>
1351 Optional. Compute GRE checksums on outgoing packets. Default is
1352 disabled, set to <code>true</code> to enable. Checksums present on
1353 incoming packets will be validated regardless of this setting.
1357 GRE checksums impose a significant performance penalty because they
1358 cover the entire packet. The encapsulated L3, L4, and L7 packet
1359 contents typically have their own checksums, so this additional
1360 checksum only adds value for the GRE and encapsulated L2 headers.
1364 This option is supported for <code>ipsec_gre</code>, but not useful
1365 because GRE checksums are weaker than, and redundant with, IPsec
1366 payload authentication.
1371 <group title="Tunnel Options: ipsec_gre only">
1373 Only <code>ipsec_gre</code> interfaces support these options.
1376 <column name="options" key="peer_cert">
1377 Required for certificate authentication. A string containing the
1378 peer's certificate in PEM format. Additionally the host's
1379 certificate must be specified with the <code>certificate</code>
1383 <column name="options" key="certificate">
1384 Required for certificate authentication. The name of a PEM file
1385 containing a certificate that will be presented to the peer during
1389 <column name="options" key="private_key">
1390 Optional for certificate authentication. The name of a PEM file
1391 containing the private key associated with <code>certificate</code>.
1392 If <code>certificate</code> contains the private key, this option may
1396 <column name="options" key="psk">
1397 Required for pre-shared key authentication. Specifies a pre-shared
1398 key for authentication that must be identical on both sides of the
1404 <group title="Patch Options">
1406 Only <code>patch</code> interfaces support these options.
1409 <column name="options" key="peer">
1410 The <ref column="name"/> of the <ref table="Interface"/> for the other
1411 side of the patch. The named <ref table="Interface"/>'s own
1412 <code>peer</code> option must specify this <ref table="Interface"/>'s
1413 name. That is, the two patch interfaces must have reversed <ref
1414 column="name"/> and <code>peer</code> values.
1418 <group title="Interface Status">
1420 Status information about interfaces attached to bridges, updated every
1421 5 seconds. Not all interfaces have all of these properties; virtual
1422 interfaces don't have a link speed, for example. Non-applicable
1423 columns will have empty values.
1425 <column name="admin_state">
1427 The administrative state of the physical network link.
1431 <column name="link_state">
1433 The observed state of the physical network link. This is ordinarily
1434 the link's carrier status. If the interface's <ref table="Port"/> is
1435 a bond configured for miimon monitoring, it is instead the network
1436 link's miimon status.
1440 <column name="link_resets">
1442 The number of times Open vSwitch has observed the
1443 <ref column="link_state"/> of this <ref table="Interface"/> change.
1447 <column name="link_speed">
1449 The negotiated speed of the physical network link.
1450 Valid values are positive integers greater than 0.
1454 <column name="duplex">
1456 The duplex mode of the physical network link.
1462 The MTU (maximum transmission unit); i.e. the largest
1463 amount of data that can fit into a single Ethernet frame.
1464 The standard Ethernet MTU is 1500 bytes. Some physical media
1465 and many kinds of virtual interfaces can be configured with
1469 This column will be empty for an interface that does not
1470 have an MTU as, for example, some kinds of tunnels do not.
1474 <column name="lacp_current">
1475 Boolean value indicating LACP status for this interface. If true, this
1476 interface has current LACP information about its LACP partner. This
1477 information may be used to monitor the health of interfaces in a LACP
1478 enabled port. This column will be empty if LACP is not enabled.
1481 <column name="status">
1482 Key-value pairs that report port status. Supported status values are
1483 <ref column="type"/>-dependent; some interfaces may not have a valid
1484 <ref column="status" key="driver_name"/>, for example.
1487 <column name="status" key="driver_name">
1488 The name of the device driver controlling the network adapter.
1491 <column name="status" key="driver_version">
1492 The version string of the device driver controlling the network
1496 <column name="status" key="firmware_version">
1497 The version string of the network adapter's firmware, if available.
1500 <column name="status" key="source_ip">
1501 The source IP address used for an IPv4 tunnel end-point, such as
1502 <code>gre</code> or <code>capwap</code>.
1505 <column name="status" key="tunnel_egress_iface">
1506 Egress interface for tunnels. Currently only relevant for GRE and
1507 CAPWAP tunnels. On Linux systems, this column will show the name of
1508 the interface which is responsible for routing traffic destined for the
1509 configured <ref column="options" key="remote_ip"/>. This could be an
1510 internal interface such as a bridge port.
1513 <column name="status" key="tunnel_egress_iface_carrier"
1514 type='{"type": "string", "enum": ["set", ["down", "up"]]}'>
1515 Whether carrier is detected on <ref column="status"
1516 key="tunnel_egress_iface"/>.
1520 <group title="Statistics">
1522 Key-value pairs that report interface statistics. The current
1523 implementation updates these counters periodically. Future
1524 implementations may update them when an interface is created, when they
1525 are queried (e.g. using an OVSDB <code>select</code> operation), and
1526 just before an interface is deleted due to virtual interface hot-unplug
1527 or VM shutdown, and perhaps at other times, but not on any regular
1531 These are the same statistics reported by OpenFlow in its <code>struct
1532 ofp_port_stats</code> structure. If an interface does not support a
1533 given statistic, then that pair is omitted.
1535 <group title="Statistics: Successful transmit and receive counters">
1536 <column name="statistics" key="rx_packets">
1537 Number of received packets.
1539 <column name="statistics" key="rx_bytes">
1540 Number of received bytes.
1542 <column name="statistics" key="tx_packets">
1543 Number of transmitted packets.
1545 <column name="statistics" key="tx_bytes">
1546 Number of transmitted bytes.
1549 <group title="Statistics: Receive errors">
1550 <column name="statistics" key="rx_dropped">
1551 Number of packets dropped by RX.
1553 <column name="statistics" key="rx_frame_err">
1554 Number of frame alignment errors.
1556 <column name="statistics" key="rx_over_err">
1557 Number of packets with RX overrun.
1559 <column name="statistics" key="rx_crc_err">
1560 Number of CRC errors.
1562 <column name="statistics" key="rx_errors">
1563 Total number of receive errors, greater than or equal to the sum of
1567 <group title="Statistics: Transmit errors">
1568 <column name="statistics" key="tx_dropped">
1569 Number of packets dropped by TX.
1571 <column name="statistics" key="collisions">
1572 Number of collisions.
1574 <column name="statistics" key="tx_errors">
1575 Total number of transmit errors, greater than or equal to the sum of
1581 <group title="Ingress Policing">
1583 These settings control ingress policing for packets received on this
1584 interface. On a physical interface, this limits the rate at which
1585 traffic is allowed into the system from the outside; on a virtual
1586 interface (one connected to a virtual machine), this limits the rate at
1587 which the VM is able to transmit.
1590 Policing is a simple form of quality-of-service that simply drops
1591 packets received in excess of the configured rate. Due to its
1592 simplicity, policing is usually less accurate and less effective than
1593 egress QoS (which is configured using the <ref table="QoS"/> and <ref
1594 table="Queue"/> tables).
1597 Policing is currently implemented only on Linux. The Linux
1598 implementation uses a simple ``token bucket'' approach:
1602 The size of the bucket corresponds to <ref
1603 column="ingress_policing_burst"/>. Initially the bucket is full.
1606 Whenever a packet is received, its size (converted to tokens) is
1607 compared to the number of tokens currently in the bucket. If the
1608 required number of tokens are available, they are removed and the
1609 packet is forwarded. Otherwise, the packet is dropped.
1612 Whenever it is not full, the bucket is refilled with tokens at the
1613 rate specified by <ref column="ingress_policing_rate"/>.
1617 Policing interacts badly with some network protocols, and especially
1618 with fragmented IP packets. Suppose that there is enough network
1619 activity to keep the bucket nearly empty all the time. Then this token
1620 bucket algorithm will forward a single packet every so often, with the
1621 period depending on packet size and on the configured rate. All of the
1622 fragments of an IP packets are normally transmitted back-to-back, as a
1623 group. In such a situation, therefore, only one of these fragments
1624 will be forwarded and the rest will be dropped. IP does not provide
1625 any way for the intended recipient to ask for only the remaining
1626 fragments. In such a case there are two likely possibilities for what
1627 will happen next: either all of the fragments will eventually be
1628 retransmitted (as TCP will do), in which case the same problem will
1629 recur, or the sender will not realize that its packet has been dropped
1630 and data will simply be lost (as some UDP-based protocols will do).
1631 Either way, it is possible that no forward progress will ever occur.
1633 <column name="ingress_policing_rate">
1635 Maximum rate for data received on this interface, in kbps. Data
1636 received faster than this rate is dropped. Set to <code>0</code>
1637 (the default) to disable policing.
1641 <column name="ingress_policing_burst">
1642 <p>Maximum burst size for data received on this interface, in kb. The
1643 default burst size if set to <code>0</code> is 1000 kb. This value
1644 has no effect if <ref column="ingress_policing_rate"/>
1645 is <code>0</code>.</p>
1647 Specifying a larger burst size lets the algorithm be more forgiving,
1648 which is important for protocols like TCP that react severely to
1649 dropped packets. The burst size should be at least the size of the
1650 interface's MTU. Specifying a value that is numerically at least as
1651 large as 10% of <ref column="ingress_policing_rate"/> helps TCP come
1652 closer to achieving the full rate.
1657 <group title="Connectivity Fault Management">
1659 802.1ag Connectivity Fault Management (CFM) allows a group of
1660 Maintenance Points (MPs) called a Maintenance Association (MA) to
1661 detect connectivity problems with each other. MPs within a MA should
1662 have complete and exclusive interconnectivity. This is verified by
1663 occasionally broadcasting Continuity Check Messages (CCMs) at a
1664 configurable transmission interval.
1668 According to the 802.1ag specification, each Maintenance Point should
1669 be configured out-of-band with a list of Remote Maintenance Points it
1670 should have connectivity to. Open vSwitch differs from the
1671 specification in this area. It simply assumes the link is faulted if
1672 no Remote Maintenance Points are reachable, and considers it not
1676 <column name="cfm_mpid">
1677 A Maintenance Point ID (MPID) uniquely identifies each endpoint within
1678 a Maintenance Association. The MPID is used to identify this endpoint
1679 to other Maintenance Points in the MA. Each end of a link being
1680 monitored should have a different MPID. Must be configured to enable
1681 CFM on this <ref table="Interface"/>.
1684 <column name="cfm_fault">
1686 Indicates a connectivity fault triggered by an inability to receive
1687 heartbeats from any remote endpoint. When a fault is triggered on
1688 <ref table="Interface"/>s participating in bonds, they will be
1692 Faults can be triggered for several reasons. Most importantly they
1693 are triggered when no CCMs are received for a period of 3.5 times the
1694 transmission interval. Faults are also triggered when any CCMs
1695 indicate that a Remote Maintenance Point is not receiving CCMs but
1696 able to send them. Finally, a fault is triggered if a CCM is
1697 received which indicates unexpected configuration. Notably, this
1698 case arises when a CCM is received which advertises the local MPID.
1702 <column name="cfm_remote_mpids">
1703 When CFM is properly configured, Open vSwitch will occasionally
1704 receive CCM broadcasts. These broadcasts contain the MPID of the
1705 sending Maintenance Point. The list of MPIDs from which this
1706 <ref table="Interface"/> is receiving broadcasts from is regularly
1707 collected and written to this column.
1710 <column name="other_config" key="cfm_interval"
1711 type='{"type": "integer"}'>
1712 The interval, in milliseconds, between transmissions of CFM heartbeats.
1713 Three missed heartbeat receptions indicate a connectivity fault.
1717 <column name="other_config" key="cfm_extended"
1718 type='{"type": "boolean"}'>
1719 When <code>true</code>, the CFM module operates in extended mode. This
1720 causes it to use a nonstandard destination address to avoid conflicting
1721 with compliant implementations which may be running concurrently on the
1722 network. Furthermore, extended mode increases the accuracy of the
1723 <code>cfm_interval</code> configuration parameter by breaking wire
1724 compatibility with 802.1ag compliant implementations. Defaults to
1727 <column name="other_config" key="cfm_opstate"
1728 type='{"type": "string", "enum": ["set", ["down", "up"]]}'>
1729 When <code>down</code>, the CFM module marks all CCMs it generates as
1730 operationally down without triggering a fault. This allows remote
1731 maintenance points to choose not to forward traffic to the
1732 <ref table="Interface"/> on which this CFM module is running.
1733 Currently, in Open vSwitch, the opdown bit of CCMs affects
1734 <ref table="Interface"/>s participating in bonds, and the bundle
1735 OpenFlow action. This setting is ignored when CFM is not in extended
1736 mode. Defaults to <code>up</code>.
1739 <column name="other_config" key="cfm_ccm_vlan"
1740 type='{"type": "integer", "minInteger": 1, "maxInteger": 4095}'>
1741 When set, the CFM module will apply a VLAN tag to all CCMs it generates
1742 with the given value.
1745 <column name="other_config" key="cfm_ccm_pcp"
1746 type='{"type": "integer", "minInteger": 1, "maxInteger": 7}'>
1747 When set, the CFM module will apply a VLAN tag to all CCMs it generates
1748 with the given PCP value. The VLAN ID of the tag is governed by the
1749 value of <ref column="other_config" key="cfm_ccm_vlan"/>. If
1750 <ref column="other_config" key="cfm_ccm_vlan"/> is unset, a VLAN ID of
1756 <group title="Bonding Configuration">
1757 <column name="other_config" key="bond-stable-id"
1758 type='{"type": "integer", "minInteger": 1}'>
1759 Used in <code>stable</code> bond mode to make slave
1760 selection decisions. Allocating <ref column="other_config"
1761 key="bond-stable-id"/> values consistently across interfaces
1762 participating in a bond will guarantee consistent slave selection
1763 decisions across <code>ovs-vswitchd</code> instances when using
1764 <code>stable</code> bonding mode.
1767 <column name="other_config" key="lacp-port-id"
1768 type='{"type": "integer", "minInteger": 1, "maxInteger": 65535}'>
1769 The LACP port ID of this <ref table="Interface"/>. Port IDs are
1770 used in LACP negotiations to identify individual ports
1771 participating in a bond.
1774 <column name="other_config" key="lacp-port-priority"
1775 type='{"type": "integer", "minInteger": 1, "maxInteger": 65535}'>
1776 The LACP port priority of this <ref table="Interface"/>. In LACP
1777 negotiations <ref table="Interface"/>s with numerically lower
1778 priorities are preferred for aggregation.
1781 <column name="other_config" key="lacp-aggregation-key"
1782 type='{"type": "integer", "minInteger": 1, "maxInteger": 65535}'>
1783 The LACP aggregation key of this <ref table="Interface"/>. <ref
1784 table="Interface"/>s with different aggregation keys may not be active
1785 within a given <ref table="Port"/> at the same time.
1789 <group title="Virtual Machine Identifiers">
1791 These key-value pairs specifically apply to an interface that
1792 represents a virtual Ethernet interface connected to a virtual
1793 machine. These key-value pairs should not be present for other types
1794 of interfaces. Keys whose names end in <code>-uuid</code> have
1795 values that uniquely identify the entity in question. For a Citrix
1796 XenServer hypervisor, these values are UUIDs in RFC 4122 format.
1797 Other hypervisors may use other formats.
1800 <column name="external_ids" key="attached-mac">
1801 The MAC address programmed into the ``virtual hardware'' for this
1802 interface, in the form
1803 <var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>.
1804 For Citrix XenServer, this is the value of the <code>MAC</code> field
1805 in the VIF record for this interface.
1808 <column name="external_ids" key="iface-id">
1809 A system-unique identifier for the interface. On XenServer, this will
1810 commonly be the same as <ref column="external_ids" key="xs-vif-uuid"/>.
1813 <column name="external_ids" key="xs-vif-uuid">
1814 The virtual interface associated with this interface.
1817 <column name="external_ids" key="xs-network-uuid">
1818 The virtual network to which this interface is attached.
1821 <column name="external_ids" key="xs-vm-uuid">
1822 The VM to which this interface belongs.
1826 <group title="VLAN Splinters">
1828 The ``VLAN splinters'' feature increases Open vSwitch compatibility
1829 with buggy network drivers in old versions of Linux that do not
1830 properly support VLANs when VLAN devices are not used, at some cost
1831 in memory and performance.
1835 When VLAN splinters are enabled on a particular interface, Open vSwitch
1836 creates a VLAN device for each in-use VLAN. For sending traffic tagged
1837 with a VLAN on the interface, it substitutes the VLAN device. Traffic
1838 received on the VLAN device is treated as if it had been received on
1839 the interface on the particular VLAN.
1843 VLAN splinters consider a VLAN to be in use if:
1848 The VLAN is the <ref table="Port" column="tag"/> value in any <ref
1849 table="Port"/> record.
1853 The VLAN is listed within the <ref table="Port" column="trunks"/>
1854 column of the <ref table="Port"/> record of an interface on which
1855 VLAN splinters are enabled.
1857 An empty <ref table="Port" column="trunks"/> does not influence the
1858 in-use VLANs: creating 4,096 VLAN devices is impractical because it
1859 will exceed the current 1,024 port per datapath limit.
1863 An OpenFlow flow within any bridge matches the VLAN.
1868 The same set of in-use VLANs applies to every interface on which VLAN
1869 splinters are enabled. That is, the set is not chosen separately for
1870 each interface but selected once as the union of all in-use VLANs based
1875 It does not make sense to enable VLAN splinters on an interface for an
1876 access port, or on an interface that is not a physical port.
1880 VLAN splinters are deprecated. When broken device drivers are no
1881 longer in widespread use, we will delete this feature.
1884 <column name="other_config" key="enable-vlan-splinters"
1885 type='{"type": "boolean"}'>
1887 Set to <code>true</code> to enable VLAN splinters on this interface.
1888 Defaults to <code>false</code>.
1892 VLAN splinters increase kernel and userspace memory overhead, so do
1893 not use them unless they are needed.
1898 <group title="Common Columns">
1899 The overall purpose of these columns is described under <code>Common
1900 Columns</code> at the beginning of this document.
1902 <column name="other_config"/>
1903 <column name="external_ids"/>
1907 <table name="Flow_Table" title="OpenFlow table configuration">
1908 <p>Configuration for a particular OpenFlow table.</p>
1910 <column name="name">
1911 The table's name. Set this column to change the name that controllers
1912 will receive when they request table statistics, e.g. <code>ovs-ofctl
1913 dump-tables</code>. The name does not affect switch behavior.
1916 <column name="flow_limit">
1917 If set, limits the number of flows that may be added to the table. Open
1918 vSwitch may limit the number of flows in a table for other reasons,
1919 e.g. due to hardware limitations or for resource availability or
1920 performance reasons.
1923 <column name="overflow_policy">
1925 Controls the switch's behavior when an OpenFlow flow table modification
1926 request would add flows in excess of <ref column="flow_limit"/>. The
1927 supported values are:
1931 <dt><code>refuse</code></dt>
1933 Refuse to add the flow or flows. This is also the default policy
1934 when <ref column="overflow_policy"/> is unset.
1937 <dt><code>evict</code></dt>
1939 Delete the flow that will expire soonest. See <ref column="groups"/>
1945 <column name="groups">
1947 When <ref column="overflow_policy"/> is <code>evict</code>, this
1948 controls how flows are chosen for eviction when the flow table would
1949 otherwise exceed <ref column="flow_limit"/> flows. Its value is a set
1950 of NXM fields or sub-fields, each of which takes one of the forms
1951 <code><var>field</var>[]</code> or
1952 <code><var>field</var>[<var>start</var>..<var>end</var>]</code>,
1953 e.g. <code>NXM_OF_IN_PORT[]</code>. Please see
1954 <code>nicira-ext.h</code> for a complete list of NXM field names.
1958 When a flow must be evicted due to overflow, the flow to evict is
1959 chosen through an approximation of the following algorithm:
1964 Divide the flows in the table into groups based on the values of the
1965 specified fields or subfields, so that all of the flows in a given
1966 group have the same values for those fields. If a flow does not
1967 specify a given field, that field's value is treated as 0.
1971 Consider the flows in the largest group, that is, the group that
1972 contains the greatest number of flows. If two or more groups all
1973 have the same largest number of flows, consider the flows in all of
1978 Among the flows under consideration, choose the flow that expires
1979 soonest for eviction.
1984 The eviction process only considers flows that have an idle timeout or
1985 a hard timeout. That is, eviction never deletes permanent flows.
1986 (Permanent flows do count against <ref column="flow_limit"/>.
1990 Open vSwitch ignores any invalid or unknown field specifications.
1994 When <ref column="overflow_policy"/> is not <code>evict</code>, this
1995 column has no effect.
2000 <table name="QoS" title="Quality of Service configuration">
2001 <p>Quality of Service (QoS) configuration for each Port that
2004 <column name="type">
2005 <p>The type of QoS to implement. The <ref table="Open_vSwitch"
2006 column="capabilities"/> column in the <ref table="Open_vSwitch"/> table
2007 identifies the types that a switch actually supports. The currently
2008 defined types are listed below:</p>
2010 <dt><code>linux-htb</code></dt>
2012 Linux ``hierarchy token bucket'' classifier. See tc-htb(8) (also at
2013 <code>http://linux.die.net/man/8/tc-htb</code>) and the HTB manual
2014 (<code>http://luxik.cdi.cz/~devik/qos/htb/manual/userg.htm</code>)
2015 for information on how this classifier works and how to configure it.
2019 <dt><code>linux-hfsc</code></dt>
2021 Linux "Hierarchical Fair Service Curve" classifier.
2022 See <code>http://linux-ip.net/articles/hfsc.en/</code> for
2023 information on how this classifier works.
2028 <column name="queues">
2029 <p>A map from queue numbers to <ref table="Queue"/> records. The
2030 supported range of queue numbers depend on <ref column="type"/>. The
2031 queue numbers are the same as the <code>queue_id</code> used in
2032 OpenFlow in <code>struct ofp_action_enqueue</code> and other
2033 structures. Queue 0 is used by OpenFlow output actions that do not
2034 specify a specific queue.</p>
2037 <group title="Configuration for linux-htb and linux-hfsc">
2039 The <code>linux-htb</code> and <code>linux-hfsc</code> classes support
2040 the following key-value pair:
2043 <column name="other_config" key="max-rate" type='{"type": "integer"}'>
2044 Maximum rate shared by all queued traffic, in bit/s. Optional. If not
2045 specified, for physical interfaces, the default is the link rate. For
2046 other interfaces or if the link rate cannot be determined, the default
2047 is currently 100 Mbps.
2051 <group title="Common Columns">
2052 The overall purpose of these columns is described under <code>Common
2053 Columns</code> at the beginning of this document.
2055 <column name="other_config"/>
2056 <column name="external_ids"/>
2060 <table name="Queue" title="QoS output queue.">
2061 <p>A configuration for a port output queue, used in configuring Quality of
2062 Service (QoS) features. May be referenced by <ref column="queues"
2063 table="QoS"/> column in <ref table="QoS"/> table.</p>
2065 <column name="dscp">
2066 If set, Open vSwitch will mark all traffic egressing this
2067 <ref table="Queue"/> with the given DSCP bits. Traffic egressing the
2068 default <ref table="Queue"/> is only marked if it was explicitly selected
2069 as the <ref table="Queue"/> at the time the packet was output. If unset,
2070 the DSCP bits of traffic egressing this <ref table="Queue"/> will remain
2074 <group title="Configuration for linux-htb QoS">
2076 <ref table="QoS"/> <ref table="QoS" column="type"/>
2077 <code>linux-htb</code> may use <code>queue_id</code>s less than 61440.
2078 It has the following key-value pairs defined.
2081 <column name="other_config" key="min-rate"
2082 type='{"type": "integer", "minInteger": 1}'>
2083 Minimum guaranteed bandwidth, in bit/s.
2086 <column name="other_config" key="max-rate"
2087 type='{"type": "integer", "minInteger": 1}'>
2088 Maximum allowed bandwidth, in bit/s. Optional. If specified, the
2089 queue's rate will not be allowed to exceed the specified value, even
2090 if excess bandwidth is available. If unspecified, defaults to no
2094 <column name="other_config" key="burst"
2095 type='{"type": "integer", "minInteger": 1}'>
2096 Burst size, in bits. This is the maximum amount of ``credits'' that a
2097 queue can accumulate while it is idle. Optional. Details of the
2098 <code>linux-htb</code> implementation require a minimum burst size, so
2099 a too-small <code>burst</code> will be silently ignored.
2102 <column name="other_config" key="priority"
2103 type='{"type": "integer", "minInteger": 0, "maxInteger": 4294967295}'>
2104 A queue with a smaller <code>priority</code> will receive all the
2105 excess bandwidth that it can use before a queue with a larger value
2106 receives any. Specific priority values are unimportant; only relative
2107 ordering matters. Defaults to 0 if unspecified.
2111 <group title="Configuration for linux-hfsc QoS">
2113 <ref table="QoS"/> <ref table="QoS" column="type"/>
2114 <code>linux-hfsc</code> may use <code>queue_id</code>s less than 61440.
2115 It has the following key-value pairs defined.
2118 <column name="other_config" key="min-rate"
2119 type='{"type": "integer", "minInteger": 1}'>
2120 Minimum guaranteed bandwidth, in bit/s.
2123 <column name="other_config" key="max-rate"
2124 type='{"type": "integer", "minInteger": 1}'>
2125 Maximum allowed bandwidth, in bit/s. Optional. If specified, the
2126 queue's rate will not be allowed to exceed the specified value, even if
2127 excess bandwidth is available. If unspecified, defaults to no
2132 <group title="Common Columns">
2133 The overall purpose of these columns is described under <code>Common
2134 Columns</code> at the beginning of this document.
2136 <column name="other_config"/>
2137 <column name="external_ids"/>
2141 <table name="Mirror" title="Port mirroring.">
2142 <p>A port mirror within a <ref table="Bridge"/>.</p>
2143 <p>A port mirror configures a bridge to send selected frames to special
2144 ``mirrored'' ports, in addition to their normal destinations. Mirroring
2145 traffic may also be referred to as SPAN or RSPAN, depending on how
2146 the mirrored traffic is sent.</p>
2148 <column name="name">
2149 Arbitrary identifier for the <ref table="Mirror"/>.
2152 <group title="Selecting Packets for Mirroring">
2154 To be selected for mirroring, a given packet must enter or leave the
2155 bridge through a selected port and it must also be in one of the
2159 <column name="select_all">
2160 If true, every packet arriving or departing on any port is
2161 selected for mirroring.
2164 <column name="select_dst_port">
2165 Ports on which departing packets are selected for mirroring.
2168 <column name="select_src_port">
2169 Ports on which arriving packets are selected for mirroring.
2172 <column name="select_vlan">
2173 VLANs on which packets are selected for mirroring. An empty set
2174 selects packets on all VLANs.
2178 <group title="Mirroring Destination Configuration">
2180 These columns are mutually exclusive. Exactly one of them must be
2184 <column name="output_port">
2185 <p>Output port for selected packets, if nonempty.</p>
2186 <p>Specifying a port for mirror output reserves that port exclusively
2187 for mirroring. No frames other than those selected for mirroring
2189 will be forwarded to the port, and any frames received on the port
2190 will be discarded.</p>
2192 The output port may be any kind of port supported by Open vSwitch.
2193 It may be, for example, a physical port (sometimes called SPAN) or a
2198 <column name="output_vlan">
2199 <p>Output VLAN for selected packets, if nonempty.</p>
2200 <p>The frames will be sent out all ports that trunk
2201 <ref column="output_vlan"/>, as well as any ports with implicit VLAN
2202 <ref column="output_vlan"/>. When a mirrored frame is sent out a
2203 trunk port, the frame's VLAN tag will be set to
2204 <ref column="output_vlan"/>, replacing any existing tag; when it is
2205 sent out an implicit VLAN port, the frame will not be tagged. This
2206 type of mirroring is sometimes called RSPAN.</p>
2208 The following destination MAC addresses will not be mirrored to a
2209 VLAN to avoid confusing switches that interpret the protocols that
2213 <dt><code>01:80:c2:00:00:00</code></dt>
2214 <dd>IEEE 802.1D Spanning Tree Protocol (STP).</dd>
2216 <dt><code>01:80:c2:00:00:01</code></dt>
2217 <dd>IEEE Pause frame.</dd>
2219 <dt><code>01:80:c2:00:00:0<var>x</var></code></dt>
2220 <dd>Other reserved protocols.</dd>
2222 <dt><code>01:00:0c:cc:cc:cc</code></dt>
2224 Cisco Discovery Protocol (CDP), VLAN Trunking Protocol (VTP),
2225 Dynamic Trunking Protocol (DTP), Port Aggregation Protocol (PAgP),
2229 <dt><code>01:00:0c:cc:cc:cd</code></dt>
2230 <dd>Cisco Shared Spanning Tree Protocol PVSTP+.</dd>
2232 <dt><code>01:00:0c:cd:cd:cd</code></dt>
2233 <dd>Cisco STP Uplink Fast.</dd>
2235 <dt><code>01:00:0c:00:00:00</code></dt>
2236 <dd>Cisco Inter Switch Link.</dd>
2238 <p><em>Please note:</em> Mirroring to a VLAN can disrupt a network that
2239 contains unmanaged switches. Consider an unmanaged physical switch
2240 with two ports: port 1, connected to an end host, and port 2,
2241 connected to an Open vSwitch configured to mirror received packets
2242 into VLAN 123 on port 2. Suppose that the end host sends a packet on
2243 port 1 that the physical switch forwards to port 2. The Open vSwitch
2244 forwards this packet to its destination and then reflects it back on
2245 port 2 in VLAN 123. This reflected packet causes the unmanaged
2246 physical switch to replace the MAC learning table entry, which
2247 correctly pointed to port 1, with one that incorrectly points to port
2248 2. Afterward, the physical switch will direct packets destined for
2249 the end host to the Open vSwitch on port 2, instead of to the end
2250 host on port 1, disrupting connectivity. If mirroring to a VLAN is
2251 desired in this scenario, then the physical switch must be replaced
2252 by one that learns Ethernet addresses on a per-VLAN basis. In
2253 addition, learning should be disabled on the VLAN containing mirrored
2254 traffic. If this is not done then intermediate switches will learn
2255 the MAC address of each end host from the mirrored traffic. If
2256 packets being sent to that end host are also mirrored, then they will
2257 be dropped since the switch will attempt to send them out the input
2258 port. Disabling learning for the VLAN will cause the switch to
2259 correctly send the packet out all ports configured for that VLAN. If
2260 Open vSwitch is being used as an intermediate switch, learning can be
2261 disabled by adding the mirrored VLAN to <ref column="flood_vlans"/>
2262 in the appropriate <ref table="Bridge"/> table or tables.</p>
2264 Mirroring to a GRE tunnel has fewer caveats than mirroring to a
2265 VLAN and should generally be preferred.
2270 <group title="Statistics: Mirror counters">
2272 Key-value pairs that report mirror statistics.
2274 <column name="statistics" key="tx_packets">
2275 Number of packets transmitted through this mirror.
2277 <column name="statistics" key="tx_bytes">
2278 Number of bytes transmitted through this mirror.
2282 <group title="Common Columns">
2283 The overall purpose of these columns is described under <code>Common
2284 Columns</code> at the beginning of this document.
2286 <column name="external_ids"/>
2290 <table name="Controller" title="OpenFlow controller configuration.">
2291 <p>An OpenFlow controller.</p>
2294 Open vSwitch supports two kinds of OpenFlow controllers:
2298 <dt>Primary controllers</dt>
2301 This is the kind of controller envisioned by the OpenFlow 1.0
2302 specification. Usually, a primary controller implements a network
2303 policy by taking charge of the switch's flow table.
2307 Open vSwitch initiates and maintains persistent connections to
2308 primary controllers, retrying the connection each time it fails or
2309 drops. The <ref table="Bridge" column="fail_mode"/> column in the
2310 <ref table="Bridge"/> table applies to primary controllers.
2314 Open vSwitch permits a bridge to have any number of primary
2315 controllers. When multiple controllers are configured, Open
2316 vSwitch connects to all of them simultaneously. Because
2317 OpenFlow 1.0 does not specify how multiple controllers
2318 coordinate in interacting with a single switch, more than
2319 one primary controller should be specified only if the
2320 controllers are themselves designed to coordinate with each
2321 other. (The Nicira-defined <code>NXT_ROLE</code> OpenFlow
2322 vendor extension may be useful for this.)
2325 <dt>Service controllers</dt>
2328 These kinds of OpenFlow controller connections are intended for
2329 occasional support and maintenance use, e.g. with
2330 <code>ovs-ofctl</code>. Usually a service controller connects only
2331 briefly to inspect or modify some of a switch's state.
2335 Open vSwitch listens for incoming connections from service
2336 controllers. The service controllers initiate and, if necessary,
2337 maintain the connections from their end. The <ref table="Bridge"
2338 column="fail_mode"/> column in the <ref table="Bridge"/> table does
2339 not apply to service controllers.
2343 Open vSwitch supports configuring any number of service controllers.
2349 The <ref column="target"/> determines the type of controller.
2352 <group title="Core Features">
2353 <column name="target">
2354 <p>Connection method for controller.</p>
2356 The following connection methods are currently supported for primary
2360 <dt><code>ssl:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2362 <p>The specified SSL <var>port</var> (default: 6633) on the host at
2363 the given <var>ip</var>, which must be expressed as an IP address
2364 (not a DNS name). The <ref table="Open_vSwitch" column="ssl"/>
2365 column in the <ref table="Open_vSwitch"/> table must point to a
2366 valid SSL configuration when this form is used.</p>
2367 <p>SSL support is an optional feature that is not always built as
2368 part of Open vSwitch.</p>
2370 <dt><code>tcp:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2371 <dd>The specified TCP <var>port</var> (default: 6633) on the host at
2372 the given <var>ip</var>, which must be expressed as an IP address
2373 (not a DNS name).</dd>
2376 The following connection methods are currently supported for service
2380 <dt><code>pssl:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2383 Listens for SSL connections on the specified TCP <var>port</var>
2384 (default: 6633). If <var>ip</var>, which must be expressed as an
2385 IP address (not a DNS name), is specified, then connections are
2386 restricted to the specified local IP address.
2389 The <ref table="Open_vSwitch" column="ssl"/> column in the <ref
2390 table="Open_vSwitch"/> table must point to a valid SSL
2391 configuration when this form is used.
2393 <p>SSL support is an optional feature that is not always built as
2394 part of Open vSwitch.</p>
2396 <dt><code>ptcp:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2398 Listens for connections on the specified TCP <var>port</var>
2399 (default: 6633). If <var>ip</var>, which must be expressed as an
2400 IP address (not a DNS name), is specified, then connections are
2401 restricted to the specified local IP address.
2404 <p>When multiple controllers are configured for a single bridge, the
2405 <ref column="target"/> values must be unique. Duplicate
2406 <ref column="target"/> values yield unspecified results.</p>
2409 <column name="connection_mode">
2410 <p>If it is specified, this setting must be one of the following
2411 strings that describes how Open vSwitch contacts this OpenFlow
2412 controller over the network:</p>
2415 <dt><code>in-band</code></dt>
2416 <dd>In this mode, this controller's OpenFlow traffic travels over the
2417 bridge associated with the controller. With this setting, Open
2418 vSwitch allows traffic to and from the controller regardless of the
2419 contents of the OpenFlow flow table. (Otherwise, Open vSwitch
2420 would never be able to connect to the controller, because it did
2421 not have a flow to enable it.) This is the most common connection
2422 mode because it is not necessary to maintain two independent
2424 <dt><code>out-of-band</code></dt>
2425 <dd>In this mode, OpenFlow traffic uses a control network separate
2426 from the bridge associated with this controller, that is, the
2427 bridge does not use any of its own network devices to communicate
2428 with the controller. The control network must be configured
2429 separately, before or after <code>ovs-vswitchd</code> is started.
2433 <p>If not specified, the default is implementation-specific.</p>
2437 <group title="Controller Failure Detection and Handling">
2438 <column name="max_backoff">
2439 Maximum number of milliseconds to wait between connection attempts.
2440 Default is implementation-specific.
2443 <column name="inactivity_probe">
2444 Maximum number of milliseconds of idle time on connection to
2445 controller before sending an inactivity probe message. If Open
2446 vSwitch does not communicate with the controller for the specified
2447 number of seconds, it will send a probe. If a response is not
2448 received for the same additional amount of time, Open vSwitch
2449 assumes the connection has been broken and attempts to reconnect.
2450 Default is implementation-specific. A value of 0 disables
2455 <group title="OpenFlow Rate Limiting">
2456 <column name="controller_rate_limit">
2458 The maximum rate at which the switch will forward packets to the
2459 OpenFlow controller, in packets per second. This feature prevents a
2460 single bridge from overwhelming the controller. If not specified,
2461 the default is implementation-specific.
2465 In addition, when a high rate triggers rate-limiting, Open vSwitch
2466 queues controller packets for each port and transmits them to the
2467 controller at the configured rate. The <ref
2468 column="controller_burst_limit"/> value limits the number of queued
2469 packets. Ports on a bridge share the packet queue fairly.
2473 Open vSwitch maintains two such packet rate-limiters per bridge: one
2474 for packets sent up to the controller because they do not correspond
2475 to any flow, and the other for packets sent up to the controller by
2476 request through flow actions. When both rate-limiters are filled with
2477 packets, the actual rate that packets are sent to the controller is
2478 up to twice the specified rate.
2482 <column name="controller_burst_limit">
2483 In conjunction with <ref column="controller_rate_limit"/>,
2484 the maximum number of unused packet credits that the bridge will
2485 allow to accumulate, in packets. If not specified, the default
2486 is implementation-specific.
2490 <group title="Additional In-Band Configuration">
2491 <p>These values are considered only in in-band control mode (see
2492 <ref column="connection_mode"/>).</p>
2494 <p>When multiple controllers are configured on a single bridge, there
2495 should be only one set of unique values in these columns. If different
2496 values are set for these columns in different controllers, the effect
2499 <column name="local_ip">
2500 The IP address to configure on the local port,
2501 e.g. <code>192.168.0.123</code>. If this value is unset, then
2502 <ref column="local_netmask"/> and <ref column="local_gateway"/> are
2506 <column name="local_netmask">
2507 The IP netmask to configure on the local port,
2508 e.g. <code>255.255.255.0</code>. If <ref column="local_ip"/> is set
2509 but this value is unset, then the default is chosen based on whether
2510 the IP address is class A, B, or C.
2513 <column name="local_gateway">
2514 The IP address of the gateway to configure on the local port, as a
2515 string, e.g. <code>192.168.0.1</code>. Leave this column unset if
2516 this network has no gateway.
2520 <group title="Controller Status">
2521 <column name="is_connected">
2522 <code>true</code> if currently connected to this controller,
2523 <code>false</code> otherwise.
2527 type='{"type": "string", "enum": ["set", ["other", "master", "slave"]]}'>
2528 <p>The level of authority this controller has on the associated
2529 bridge. Possible values are:</p>
2531 <dt><code>other</code></dt>
2532 <dd>Allows the controller access to all OpenFlow features.</dd>
2533 <dt><code>master</code></dt>
2534 <dd>Equivalent to <code>other</code>, except that there may be at
2535 most one master controller at a time. When a controller configures
2536 itself as <code>master</code>, any existing master is demoted to
2537 the <code>slave</code>role.</dd>
2538 <dt><code>slave</code></dt>
2539 <dd>Allows the controller read-only access to OpenFlow features.
2540 Attempts to modify the flow table will be rejected with an
2541 error. Slave controllers do not receive OFPT_PACKET_IN or
2542 OFPT_FLOW_REMOVED messages, but they do receive OFPT_PORT_STATUS
2547 <column name="status" key="last_error">
2548 A human-readable description of the last error on the connection
2549 to the controller; i.e. <code>strerror(errno)</code>. This key
2550 will exist only if an error has occurred.
2553 <column name="status" key="state"
2554 type='{"type": "string", "enum": ["set", ["VOID", "BACKOFF", "CONNECTING", "ACTIVE", "IDLE"]]}'>
2556 The state of the connection to the controller:
2559 <dt><code>VOID</code></dt>
2560 <dd>Connection is disabled.</dd>
2562 <dt><code>BACKOFF</code></dt>
2563 <dd>Attempting to reconnect at an increasing period.</dd>
2565 <dt><code>CONNECTING</code></dt>
2566 <dd>Attempting to connect.</dd>
2568 <dt><code>ACTIVE</code></dt>
2569 <dd>Connected, remote host responsive.</dd>
2571 <dt><code>IDLE</code></dt>
2572 <dd>Connection is idle. Waiting for response to keep-alive.</dd>
2575 These values may change in the future. They are provided only for
2580 <column name="status" key="sec_since_connect"
2581 type='{"type": "integer", "minInteger": 0}'>
2582 The amount of time since this controller last successfully connected to
2583 the switch (in seconds). Value is empty if controller has never
2584 successfully connected.
2587 <column name="status" key="sec_since_disconnect"
2588 type='{"type": "integer", "minInteger": 1}'>
2589 The amount of time since this controller last disconnected from
2590 the switch (in seconds). Value is empty if controller has never
2595 <group title="Common Columns">
2596 The overall purpose of these columns is described under <code>Common
2597 Columns</code> at the beginning of this document.
2599 <column name="external_ids"/>
2603 <table name="Manager" title="OVSDB management connection.">
2605 Configuration for a database connection to an Open vSwitch database
2610 This table primarily configures the Open vSwitch database
2611 (<code>ovsdb-server</code>), not the Open vSwitch switch
2612 (<code>ovs-vswitchd</code>). The switch does read the table to determine
2613 what connections should be treated as in-band.
2617 The Open vSwitch database server can initiate and maintain active
2618 connections to remote clients. It can also listen for database
2622 <group title="Core Features">
2623 <column name="target">
2624 <p>Connection method for managers.</p>
2626 The following connection methods are currently supported:
2629 <dt><code>ssl:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2632 The specified SSL <var>port</var> (default: 6632) on the host at
2633 the given <var>ip</var>, which must be expressed as an IP address
2634 (not a DNS name). The <ref table="Open_vSwitch" column="ssl"/>
2635 column in the <ref table="Open_vSwitch"/> table must point to a
2636 valid SSL configuration when this form is used.
2639 SSL support is an optional feature that is not always built as
2640 part of Open vSwitch.
2644 <dt><code>tcp:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2646 The specified TCP <var>port</var> (default: 6632) on the host at
2647 the given <var>ip</var>, which must be expressed as an IP address
2650 <dt><code>pssl:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2653 Listens for SSL connections on the specified TCP <var>port</var>
2654 (default: 6632). If <var>ip</var>, which must be expressed as an
2655 IP address (not a DNS name), is specified, then connections are
2656 restricted to the specified local IP address.
2659 The <ref table="Open_vSwitch" column="ssl"/> column in the <ref
2660 table="Open_vSwitch"/> table must point to a valid SSL
2661 configuration when this form is used.
2664 SSL support is an optional feature that is not always built as
2665 part of Open vSwitch.
2668 <dt><code>ptcp:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2670 Listens for connections on the specified TCP <var>port</var>
2671 (default: 6632). If <var>ip</var>, which must be expressed as an
2672 IP address (not a DNS name), is specified, then connections are
2673 restricted to the specified local IP address.
2676 <p>When multiple managers are configured, the <ref column="target"/>
2677 values must be unique. Duplicate <ref column="target"/> values yield
2678 unspecified results.</p>
2681 <column name="connection_mode">
2683 If it is specified, this setting must be one of the following strings
2684 that describes how Open vSwitch contacts this OVSDB client over the
2689 <dt><code>in-band</code></dt>
2691 In this mode, this connection's traffic travels over a bridge
2692 managed by Open vSwitch. With this setting, Open vSwitch allows
2693 traffic to and from the client regardless of the contents of the
2694 OpenFlow flow table. (Otherwise, Open vSwitch would never be able
2695 to connect to the client, because it did not have a flow to enable
2696 it.) This is the most common connection mode because it is not
2697 necessary to maintain two independent networks.
2699 <dt><code>out-of-band</code></dt>
2701 In this mode, the client's traffic uses a control network separate
2702 from that managed by Open vSwitch, that is, Open vSwitch does not
2703 use any of its own network devices to communicate with the client.
2704 The control network must be configured separately, before or after
2705 <code>ovs-vswitchd</code> is started.
2710 If not specified, the default is implementation-specific.
2715 <group title="Client Failure Detection and Handling">
2716 <column name="max_backoff">
2717 Maximum number of milliseconds to wait between connection attempts.
2718 Default is implementation-specific.
2721 <column name="inactivity_probe">
2722 Maximum number of milliseconds of idle time on connection to the client
2723 before sending an inactivity probe message. If Open vSwitch does not
2724 communicate with the client for the specified number of seconds, it
2725 will send a probe. If a response is not received for the same
2726 additional amount of time, Open vSwitch assumes the connection has been
2727 broken and attempts to reconnect. Default is implementation-specific.
2728 A value of 0 disables inactivity probes.
2732 <group title="Status">
2733 <column name="is_connected">
2734 <code>true</code> if currently connected to this manager,
2735 <code>false</code> otherwise.
2738 <column name="status" key="last_error">
2739 A human-readable description of the last error on the connection
2740 to the manager; i.e. <code>strerror(errno)</code>. This key
2741 will exist only if an error has occurred.
2744 <column name="status" key="state"
2745 type='{"type": "string", "enum": ["set", ["VOID", "BACKOFF", "CONNECTING", "ACTIVE", "IDLE"]]}'>
2747 The state of the connection to the manager:
2750 <dt><code>VOID</code></dt>
2751 <dd>Connection is disabled.</dd>
2753 <dt><code>BACKOFF</code></dt>
2754 <dd>Attempting to reconnect at an increasing period.</dd>
2756 <dt><code>CONNECTING</code></dt>
2757 <dd>Attempting to connect.</dd>
2759 <dt><code>ACTIVE</code></dt>
2760 <dd>Connected, remote host responsive.</dd>
2762 <dt><code>IDLE</code></dt>
2763 <dd>Connection is idle. Waiting for response to keep-alive.</dd>
2766 These values may change in the future. They are provided only for
2771 <column name="status" key="sec_since_connect"
2772 type='{"type": "integer", "minInteger": 0}'>
2773 The amount of time since this manager last successfully connected
2774 to the database (in seconds). Value is empty if manager has never
2775 successfully connected.
2778 <column name="status" key="sec_since_disconnect"
2779 type='{"type": "integer", "minInteger": 0}'>
2780 The amount of time since this manager last disconnected from the
2781 database (in seconds). Value is empty if manager has never
2785 <column name="status" key="locks_held">
2786 Space-separated list of the names of OVSDB locks that the connection
2787 holds. Omitted if the connection does not hold any locks.
2790 <column name="status" key="locks_waiting">
2791 Space-separated list of the names of OVSDB locks that the connection is
2792 currently waiting to acquire. Omitted if the connection is not waiting
2796 <column name="status" key="locks_lost">
2797 Space-separated list of the names of OVSDB locks that the connection
2798 has had stolen by another OVSDB client. Omitted if no locks have been
2799 stolen from this connection.
2802 <column name="status" key="n_connections"
2803 type='{"type": "integer", "minInteger": 2}'>
2805 When <ref column="target"/> specifies a connection method that
2806 listens for inbound connections (e.g. <code>ptcp:</code> or
2807 <code>pssl:</code>) and more than one connection is actually active,
2808 the value is the number of active connections. Otherwise, this
2809 key-value pair is omitted.
2812 When multiple connections are active, status columns and key-value
2813 pairs (other than this one) report the status of one arbitrarily
2819 <group title="Common Columns">
2820 The overall purpose of these columns is described under <code>Common
2821 Columns</code> at the beginning of this document.
2823 <column name="external_ids"/>
2827 <table name="NetFlow">
2828 A NetFlow target. NetFlow is a protocol that exports a number of
2829 details about terminating IP flows, such as the principals involved
2832 <column name="targets">
2833 NetFlow targets in the form
2834 <code><var>ip</var>:<var>port</var></code>. The <var>ip</var>
2835 must be specified numerically, not as a DNS name.
2838 <column name="engine_id">
2839 Engine ID to use in NetFlow messages. Defaults to datapath index
2843 <column name="engine_type">
2844 Engine type to use in NetFlow messages. Defaults to datapath
2845 index if not specified.
2848 <column name="active_timeout">
2849 The interval at which NetFlow records are sent for flows that are
2850 still active, in seconds. A value of <code>0</code> requests the
2851 default timeout (currently 600 seconds); a value of <code>-1</code>
2852 disables active timeouts.
2855 <column name="add_id_to_interface">
2856 <p>If this column's value is <code>false</code>, the ingress and egress
2857 interface fields of NetFlow flow records are derived from OpenFlow port
2858 numbers. When it is <code>true</code>, the 7 most significant bits of
2859 these fields will be replaced by the least significant 7 bits of the
2860 engine id. This is useful because many NetFlow collectors do not
2861 expect multiple switches to be sending messages from the same host, so
2862 they do not store the engine information which could be used to
2863 disambiguate the traffic.</p>
2864 <p>When this option is enabled, a maximum of 508 ports are supported.</p>
2867 <group title="Common Columns">
2868 The overall purpose of these columns is described under <code>Common
2869 Columns</code> at the beginning of this document.
2871 <column name="external_ids"/>
2876 SSL configuration for an Open_vSwitch.
2878 <column name="private_key">
2879 Name of a PEM file containing the private key used as the switch's
2880 identity for SSL connections to the controller.
2883 <column name="certificate">
2884 Name of a PEM file containing a certificate, signed by the
2885 certificate authority (CA) used by the controller and manager,
2886 that certifies the switch's private key, identifying a trustworthy
2890 <column name="ca_cert">
2891 Name of a PEM file containing the CA certificate used to verify
2892 that the switch is connected to a trustworthy controller.
2895 <column name="bootstrap_ca_cert">
2896 If set to <code>true</code>, then Open vSwitch will attempt to
2897 obtain the CA certificate from the controller on its first SSL
2898 connection and save it to the named PEM file. If it is successful,
2899 it will immediately drop the connection and reconnect, and from then
2900 on all SSL connections must be authenticated by a certificate signed
2901 by the CA certificate thus obtained. <em>This option exposes the
2902 SSL connection to a man-in-the-middle attack obtaining the initial
2903 CA certificate.</em> It may still be useful for bootstrapping.
2906 <group title="Common Columns">
2907 The overall purpose of these columns is described under <code>Common
2908 Columns</code> at the beginning of this document.
2910 <column name="external_ids"/>
2914 <table name="sFlow">
2915 <p>An sFlow(R) target. sFlow is a protocol for remote monitoring
2918 <column name="agent">
2919 Name of the network device whose IP address should be reported as the
2920 ``agent address'' to collectors. If not specified, the agent device is
2921 figured from the first target address and the routing table. If the
2922 routing table does not contain a route to the target, the IP address
2923 defaults to the <ref table="Controller" column="local_ip"/> in the
2924 collector's <ref table="Controller"/>. If an agent IP address cannot be
2925 determined any of these ways, sFlow is disabled.
2928 <column name="header">
2929 Number of bytes of a sampled packet to send to the collector.
2930 If not specified, the default is 128 bytes.
2933 <column name="polling">
2934 Polling rate in seconds to send port statistics to the collector.
2935 If not specified, defaults to 30 seconds.
2938 <column name="sampling">
2939 Rate at which packets should be sampled and sent to the collector.
2940 If not specified, defaults to 400, which means one out of 400
2941 packets, on average, will be sent to the collector.
2944 <column name="targets">
2945 sFlow targets in the form
2946 <code><var>ip</var>:<var>port</var></code>.
2949 <group title="Common Columns">
2950 The overall purpose of these columns is described under <code>Common
2951 Columns</code> at the beginning of this document.
2953 <column name="external_ids"/>
2957 <table name="Capability">
2958 <p>Records in this table describe functionality supported by the hardware
2959 and software platform on which this Open vSwitch is based. Clients
2960 should not modify this table.</p>
2962 <p>A record in this table is meaningful only if it is referenced by the
2963 <ref table="Open_vSwitch" column="capabilities"/> column in the
2964 <ref table="Open_vSwitch"/> table. The key used to reference it, called
2965 the record's ``category,'' determines the meanings of the
2966 <ref column="details"/> column. The following general forms of
2967 categories are currently defined:</p>
2970 <dt><code>qos-<var>type</var></code></dt>
2971 <dd><var>type</var> is supported as the value for
2972 <ref column="type" table="QoS"/> in the <ref table="QoS"/> table.
2976 <column name="details">
2977 <p>Key-value pairs that describe capabilities. The meaning of the pairs
2978 depends on the category key that the <ref table="Open_vSwitch"
2979 column="capabilities"/> column in the <ref table="Open_vSwitch"/> table
2980 uses to reference this record, as described above.</p>
2982 <p>The presence of a record for category <code>qos-<var>type</var></code>
2983 indicates that the switch supports <var>type</var> as the value of
2984 the <ref table="QoS" column="type"/> column in the <ref table="QoS"/>
2985 table. The following key-value pairs are defined to further describe
2986 QoS capabilities:</p>
2989 <dt><code>n-queues</code></dt>
2990 <dd>Number of supported queues, as a positive integer. Keys in the
2991 <ref table="QoS" column="queues"/> column for <ref table="QoS"/>
2992 records whose <ref table="QoS" column="type"/> value
2993 equals <var>type</var> must range between 0 and this value minus one,