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
12 ``root set'' tables noted below.
15 <table name="Open_vSwitch" title="Open vSwitch configuration.">
16 Configuration for an Open vSwitch daemon. There must be exactly
17 one record in the <ref table="Open_vSwitch"/> table.
19 <group title="Configuration">
20 <column name="bridges">
21 Set of bridges managed by the daemon.
25 SSL used globally by the daemon.
28 <column name="external_ids">
29 Key-value pairs for use by external frameworks that integrate
30 with Open vSwitch, rather than by Open vSwitch itself. System
31 integrators should either use the Open vSwitch development
32 mailing list to coordinate on common key-value definitions, or
33 choose key names that are likely to be unique. The currently
34 defined common key-value pairs are:
36 <dt><code>system-id</code></dt>
37 <dd>A unique identifier for the Open vSwitch's physical host.
38 The form of the identifier depends on the type of the host.
39 On a Citrix XenServer, this will likely be the same as
40 <code>xs-system-uuid</code>.</dd>
41 <dt><code>xs-system-uuid</code></dt>
42 <dd>The Citrix XenServer universally unique identifier for the
43 physical host as displayed by <code>xe host-list</code>.</dd>
48 <group title="Status">
49 <column name="next_cfg">
50 Sequence number for client to increment. When a client modifies
51 any part of the database configuration and wishes to wait for
52 Open vSwitch to finish applying the changes, it may increment
56 <column name="cur_cfg">
57 Sequence number that Open vSwitch sets to the current value of
58 <ref column="next_cfg"/> after it finishes applying a set of
59 configuration changes.
62 <column name="capabilities">
63 Describes functionality supported by the hardware and software platform
64 on which this Open vSwitch is based. Clients should not modify this
65 column. See the <ref table="Capability"/> description for defined
66 capability categories and the meaning of associated
67 <ref table="Capability"/> records.
70 <column name="statistics">
72 Key-value pairs that report statistics about a system running an Open
73 vSwitch. These are updated periodically (currently, every 5
74 seconds). Key-value pairs that cannot be determined or that do not
75 apply to a platform are omitted.
79 <dt><code>cpu</code></dt>
82 Number of CPU processors, threads, or cores currently online and
83 available to the operating system on which Open vSwitch is
84 running, as an integer. This may be less than the number
85 installed, if some are not online or if they are not available to
89 Open vSwitch userspace processes are not multithreaded, but the
90 Linux kernel-based datapath is.
94 <dt><code>load_average</code></dt>
97 A comma-separated list of three floating-point numbers,
98 representing the system load average over the last 1, 5, and 15
99 minutes, respectively.
103 <dt><code>memory</code></dt>
106 A comma-separated list of integers, each of which represents a
107 quantity of memory in kilobytes that describes the operating
108 system on which Open vSwitch is running. In respective order,
113 <li>Total amount of RAM allocated to the OS.</li>
114 <li>RAM allocated to the OS that is in use.</li>
115 <li>RAM that can be flushed out to disk or otherwise discarded
116 if that space is needed for another purpose. This number is
117 necessarily less than or equal to the previous value.</li>
118 <li>Total disk space allocated for swap.</li>
119 <li>Swap space currently in use.</li>
123 On Linux, all five values can be determined and are included. On
124 other operating systems, only the first two values can be
125 determined, so the list will only have two values.
129 <dt><code>process_</code><var>name</var></dt>
132 One such key-value pair will exist for each running Open vSwitch
133 daemon process, with <var>name</var> replaced by the daemon's
134 name (e.g. <code>process_ovs-vswitchd</code>). The value is a
135 comma-separated list of integers. The integers represent the
136 following, with memory measured in kilobytes and durations in
141 <li>The process's virtual memory size.</li>
142 <li>The process's resident set size.</li>
143 <li>The amount of user and system CPU time consumed by the
145 <li>The number of times that the process has crashed and been
146 automatically restarted by the monitor.</li>
147 <li>The duration since the process was started.</li>
148 <li>The duration for which the process has been running.</li>
152 The interpretation of some of these values depends on whether the
153 process was started with the <option>--monitor</option>. If it
154 was not, then the crash count will always be 0 and the two
155 durations will always be the same. If <option>--monitor</option>
156 was given, then the crash count may be positive; if it is, the
157 latter duration is the amount of time since the most recent crash
162 There will be one key-value pair for each file in Open vSwitch's
163 ``run directory'' (usually <code>/var/run/openvswitch</code>)
164 whose name ends in <code>.pid</code>, whose contents are a
165 process ID, and which is locked by a running process. The
166 <var>name</var> is taken from the pidfile's name.
170 Currently Open vSwitch is only able to obtain all of the above
171 detail on Linux systems. On other systems, the same key-value
172 pairs will be present but the values will always be the empty
177 <dt><code>file_systems</code></dt>
180 A space-separated list of information on local, writable file
181 systems. Each item in the list describes one file system and
182 consists in turn of a comma-separated list of the following:
186 <li>Mount point, e.g. <code>/</code> or <code>/var/log</code>.
187 Any spaces or commas in the mount point are replaced by
189 <li>Total size, in kilobytes, as an integer.</li>
190 <li>Amount of storage in use, in kilobytes, as an integer.</li>
194 This key-value pair is omitted if there are no local, writable
195 file systems or if Open vSwitch cannot obtain the needed
203 <group title="Version Reporting">
205 These columns report the types and versions of the hardware and
206 software running Open vSwitch. We recommend in general that software
207 should test whether specific features are supported instead of relying
208 on version number checks. These values are primarily intended for
209 reporting to human administrators.
212 <column name="ovs_version">
213 The Open vSwitch version number, e.g. <code>1.1.0</code>.
214 If Open vSwitch was configured with a build number, then it is
215 also included, e.g. <code>1.1.0+build6579</code>.
218 <column name="db_version">
220 The database schema version number in the form
221 <code><var>major</var>.<var>minor</var>.<var>tweak</var></code>,
222 e.g. <code>1.2.3</code>. Whenever the database schema is changed in
223 a non-backward compatible way (e.g. deleting a column or a table),
224 <var>major</var> is incremented. When the database schema is changed
225 in a backward compatible way (e.g. adding a new column),
226 <var>minor</var> is incremented. When the database schema is changed
227 cosmetically (e.g. reindenting its syntax), <var>tweak</var> is
232 The schema version is part of the database schema, so it can also be
233 retrieved by fetching the schema using the Open vSwitch database
238 <column name="system_type">
240 An identifier for the type of system on top of which Open vSwitch
241 runs, e.g. <code>XenServer</code> or <code>KVM</code>.
244 System integrators are responsible for choosing and setting an
245 appropriate value for this column.
249 <column name="system_version">
251 The version of the system identified by <ref column="system_type"/>,
252 e.g. <code>5.6.100-39265p</code> on XenServer 5.6.100 build 39265.
255 System integrators are responsible for choosing and setting an
256 appropriate value for this column.
262 <group title="Database Configuration">
264 These columns primarily configure the Open vSwitch database
265 (<code>ovsdb-server</code>), not the Open vSwitch switch
266 (<code>ovs-vswitchd</code>). The OVSDB database also uses the <ref
267 column="ssl"/> settings.
271 The Open vSwitch switch does read the database configuration to
272 determine remote IP addresses to which in-band control should apply.
275 <column name="manager_options">
276 Database clients to which the Open vSwitch database server should
277 connect or to which it should listen, along with options for how these
278 connection should be configured. See the <ref table="Manager"/> table
279 for more information.
284 <table name="Bridge">
286 Configuration for a bridge within an
287 <ref table="Open_vSwitch"/>.
290 A <ref table="Bridge"/> record represents an Ethernet switch with one or
291 more ``ports,'' which are the <ref table="Port"/> records pointed to by
292 the <ref table="Bridge"/>'s <ref column="ports"/> column.
295 <group title="Core Features">
297 Bridge identifier. Should be alphanumeric and no more than about 8
298 bytes long. Must be unique among the names of ports, interfaces, and
302 <column name="ports">
303 Ports included in the bridge.
306 <column name="mirrors">
307 Port mirroring configuration.
310 <column name="netflow">
311 NetFlow configuration.
314 <column name="sflow">
318 <column name="flood_vlans">
319 VLAN IDs of VLANs on which MAC address learning should be disabled, so
320 that packets are flooded instead of being sent to specific ports that
321 are believed to contain packets' destination MACs. This should
322 ordinarily be used to disable MAC learning on VLANs used for mirroring
323 (RSPAN VLANs). It may also be useful for debugging.
327 <group title="OpenFlow Configuration">
328 <column name="controller">
329 OpenFlow controller set. If unset, then no OpenFlow controllers
333 <column name="fail_mode">
334 <p>When a controller is configured, it is, ordinarily, responsible
335 for setting up all flows on the switch. Thus, if the connection to
336 the controller fails, no new network connections can be set up.
337 If the connection to the controller stays down long enough,
338 no packets can pass through the switch at all. This setting
339 determines the switch's response to such a situation. It may be set
340 to one of the following:
342 <dt><code>standalone</code></dt>
343 <dd>If no message is received from the controller for three
344 times the inactivity probe interval
345 (see <ref column="inactivity_probe"/>), then Open vSwitch
346 will take over responsibility for setting up flows. In
347 this mode, Open vSwitch causes the bridge to act like an
348 ordinary MAC-learning switch. Open vSwitch will continue
349 to retry connecting to the controller in the background
350 and, when the connection succeeds, it will discontinue its
351 standalone behavior.</dd>
352 <dt><code>secure</code></dt>
353 <dd>Open vSwitch will not set up flows on its own when the
354 controller connection fails or when no controllers are
355 defined. The bridge will continue to retry connecting to
356 any defined controllers forever.</dd>
359 <p>If this value is unset, the default is implementation-specific.</p>
360 <p>When more than one controller is configured,
361 <ref column="fail_mode"/> is considered only when none of the
362 configured controllers can be contacted.</p>
365 <column name="datapath_id">
366 Reports the OpenFlow datapath ID in use. Exactly 16 hex
367 digits. (Setting this column will have no useful effect. Set
368 <ref column="other_config"/>:<code>other-config</code>
373 <group title="Other Features">
374 <column name="datapath_type">
375 Name of datapath provider. The kernel datapath has
376 type <code>system</code>. The userspace datapath has
377 type <code>netdev</code>.
380 <column name="external_ids">
381 Key-value pairs for use by external frameworks that integrate
382 with Open vSwitch, rather than by Open vSwitch itself. System
383 integrators should either use the Open vSwitch development
384 mailing list to coordinate on common key-value definitions, or
385 choose key names that are likely to be unique. The currently
386 defined key-value pairs are:
388 <dt><code>bridge-id</code></dt>
389 <dd>A unique identifier of the bridge. On Citrix XenServer this
390 will commonly be the same as <code>xs-network-uuids</code>.</dd>
391 <dt><code>xs-network-uuids</code></dt>
392 <dd>Semicolon-delimited set of universally unique identifier(s) for
393 the network with which this bridge is associated on a Citrix
394 XenServer host. The network identifiers are RFC 4122 UUIDs as
395 displayed by, e.g., <code>xe network-list</code>.</dd>
399 <column name="other_config">
400 Key-value pairs for configuring rarely used bridge
401 features. The currently defined key-value pairs are:
403 <dt><code>datapath-id</code></dt>
405 digits to set the OpenFlow datapath ID to a specific
406 value. May not be all-zero.</dd>
407 <dt><code>disable-in-band</code></dt>
408 <dd>If set to <code>true</code>, disable in-band control on
409 the bridge regardless of controller and manager settings.</dd>
410 <dt><code>hwaddr</code></dt>
411 <dd>An Ethernet address in the form
412 <var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>
413 to set the hardware address of the local port and influence the
415 <dt><code>in-band-queue</code></dt>
417 A queue ID as a nonnegative integer. This sets the OpenFlow queue
418 ID that will be used by flows set up by in-band control on this
419 bridge. If unset, or if the port used by an in-band control flow
420 does not have QoS configured, or if the port does not have a queue
421 with the specified ID, the default queue is used instead.
428 <table name="Port" table="Port or bond configuration.">
429 <p>A port within a <ref table="Bridge"/>.</p>
430 <p>Most commonly, a port has exactly one ``interface,'' pointed to by its
431 <ref column="interfaces"/> column. Such a port logically
432 corresponds to a port on a physical Ethernet switch. A port
433 with more than one interface is a ``bonded port'' (see
434 <ref group="Bonding Configuration"/>).</p>
435 <p>Some properties that one might think as belonging to a port are actually
436 part of the port's <ref table="Interface"/> members.</p>
439 Port name. Should be alphanumeric and no more than about 8
440 bytes long. May be the same as the interface name, for
441 non-bonded ports. Must otherwise be unique among the names of
442 ports, interfaces, and bridges on a host.
445 <column name="interfaces">
446 The port's interfaces. If there is more than one, this is a
450 <group title="VLAN Configuration">
451 <p>A bridge port must be configured for VLANs in one of two
452 mutually exclusive ways:
454 <li>A ``trunk port'' has an empty value for <ref
455 column="tag"/>. Its <ref column="trunks"/> value may be
456 empty or non-empty.</li>
457 <li>An ``implicitly tagged VLAN port'' or ``access port''
458 has an nonempty value for <ref column="tag"/>. Its
459 <ref column="trunks"/> value must be empty.</li>
461 If <ref column="trunks"/> and <ref column="tag"/> are both
462 nonempty, the configuration is ill-formed.
467 If this is an access port (see above), the port's implicitly
468 tagged VLAN. Must be empty if this is a trunk port.
471 Frames arriving on trunk ports will be forwarded to this
472 port only if they are tagged with the given VLAN (or, if
473 <ref column="tag"/> is 0, then if they lack a VLAN header).
474 Frames arriving on other access ports will be forwarded to
475 this port only if they have the same <ref column="tag"/>
476 value. Frames forwarded to this port will not have an
480 When a frame with a 802.1Q header that indicates a nonzero
481 VLAN is received on an access port, it is discarded.
485 <column name="trunks">
487 If this is a trunk port (see above), the 802.1Q VLAN(s) that
488 this port trunks; if it is empty, then the port trunks all
489 VLANs. Must be empty if this is an access port.
492 Frames arriving on trunk ports are dropped if they are not
493 in one of the specified VLANs. For this purpose, packets
494 that have no VLAN header are treated as part of VLAN 0.
499 <group title="Bonding Configuration">
500 <p>A port that has more than one interface is a ``bonded port.'' Bonding
501 allows for load balancing and fail-over. Some kinds of bonding will
502 work with any kind of upstream switch:</p>
505 <dt><code>balance-slb</code></dt>
507 Balances flows among slaves based on source MAC address and output
508 VLAN, with periodic rebalancing as traffic patterns change.
511 <dt><code>active-backup</code></dt>
513 Assigns all flows to one slave, failing over to a backup slave when
514 the active slave is disabled.
519 The following modes require the upstream switch to support 802.3ad with
520 successful LACP negotiation. If LACP negotiation fails then
521 <code>balance-slb</code> style flow hashing is used as a fallback:
525 <dt><code>balance-tcp</code></dt>
527 Balances flows among slaves based on L2, L3, and L4 protocol
528 information such as destination MAC address, IP address, and TCP
534 <dt><code>stable</code></dt>
536 <p>Attempts to always assign a given flow to the same slave
537 consistently. In an effort to maintain stability, no load
538 balancing is done. Uses a similar hashing strategy to
539 <code>balance-tcp</code>, falling back to <code>balance-slb</code>
540 style hashing when LACP negotiations are unsuccessful.</p>
541 <p>Slave selection decisions are made based on
542 <code>bond-stable-id</code> if set. Otherwise, OpenFlow port
543 number is used. Decisions are consistent across all ovs-vswitchd
544 instances with equivalent <code>bond-stable-id</code>s.</p>
548 <p>These columns apply only to bonded ports. Their values are
549 otherwise ignored.</p>
551 <column name="bond_mode">
552 <p>The type of bonding used for a bonded port. Defaults to
553 <code>balance-slb</code> if unset.
557 <column name="bond_updelay">
558 <p>For a bonded port, the number of milliseconds for which carrier must
559 stay up on an interface before the interface is considered to be up.
560 Specify <code>0</code> to enable the interface immediately.</p>
561 <p>This setting is honored only when at least one bonded interface is
562 already enabled. When no interfaces are enabled, then the first bond
563 interface to come up is enabled immediately.</p>
566 <column name="bond_downdelay">
567 For a bonded port, the number of milliseconds for which carrier must
568 stay down on an interface before the interface is considered to be
569 down. Specify <code>0</code> to disable the interface immediately.
572 <column name="bond_fake_iface">
573 For a bonded port, whether to create a fake internal interface with the
574 name of the port. Use only for compatibility with legacy software that
579 <p>Configures LACP on this port. LACP allows directly connected
580 switches to negotiate which links may be bonded. LACP may be enabled
581 on non-bonded ports for the benefit of any switches they may be
582 connected to. <code>active</code> ports are allowed to initiate LACP
583 negotiations. <code>passive</code> ports are allowed to participate
584 in LACP negotiations initiated by a remote switch, but not allowed to
585 initiate such negotiations themselves. If unset Open vSwitch will
586 choose a reasonable default. </p>
591 <group title="Other Features">
593 Quality of Service configuration for this port.
597 The MAC address to use for this port for the purpose of choosing the
598 bridge's MAC address. This column does not necessarily reflect the
599 port's actual MAC address, nor will setting it change the port's actual
603 <column name="fake_bridge">
604 Does this port represent a sub-bridge for its tagged VLAN within the
605 Bridge? See ovs-vsctl(8) for more information.
608 <column name="external_ids">
610 Key-value pairs for use by external frameworks that integrate with
611 Open vSwitch, rather than by Open vSwitch itself. System integrators
612 should either use the Open vSwitch development mailing list to
613 coordinate on common key-value definitions, or choose key names that
614 are likely to be unique.
617 No key-value pairs native to <ref table="Port"/> are currently
618 defined. For fake bridges (see the <ref column="fake_bridge"/>
619 column), external IDs for the fake bridge are defined here by
620 prefixing a <ref table="Bridge"/> <ref table="Bridge"
621 column="external_ids"/> key with <code>fake-bridge-</code>,
622 e.g. <code>fake-bridge-xs-network-uuids</code>.
626 <column name="other_config">
627 Key-value pairs for configuring rarely used port features. The
628 currently defined key-value pairs are:
630 <dt><code>hwaddr</code></dt>
631 <dd>An Ethernet address in the form
632 <code><var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var></code>.</dd>
633 <dt><code>bond-rebalance-interval</code></dt>
634 <dd>For an SLB bonded port, the number of milliseconds between
635 successive attempts to rebalance the bond, that is, to
636 move source MACs and their flows from one interface on
637 the bond to another in an attempt to keep usage of each
638 interface roughly equal. The default is 10000 (10
639 seconds), and the minimum is 1000 (1 second).</dd>
640 <dt><code>bond-detect-mode</code></dt>
641 <dd> Sets the method used to detect link failures in a bonded port.
642 Options are <code>carrier</code> and <code>miimon</code>. Defaults
643 to <code>carrier</code> which uses each interface's carrier to detect
644 failures. When set to <code>miimon</code>, will check for failures
645 by polling each interface's MII. </dd>
646 <dt><code>bond-miimon-interval</code></dt>
647 <dd> The number of milliseconds between successive attempts to
648 poll each interface's MII. Only relevant on ports which use
649 <code>miimon</code> to detect failures. </dd>
650 <dt><code>bond-hash-basis</code></dt>
651 <dd> An integer hashed along with flows when choosing output slaves.
652 When changed, all flows will be assigned different hash values
653 possibly causing slave selection decisions to change.</dd>
654 <dt><code>lacp-system-id</code></dt>
655 <dd> The LACP system ID of this <ref table="Port"/>. The system ID
656 of a LACP bond is used to identify itself to its partners. Must
657 be a nonzero MAC address.</dd>
658 <dt><code>lacp-system-priority</code></dt>
659 <dd> The LACP system priority of this <ref table="Port"/>. In
660 LACP negotiations, link status decisions are made by the system
661 with the numerically lower priority. Must be a number between 1
663 <dt><code>lacp-time</code></dt>
665 <p>The LACP timing which should be used on this
666 <ref table="Port"/>. Possible values are <code>fast</code>,
667 <code>slow</code> and a positive number of milliseconds. By
668 default <code>slow</code> is used. When configured to be
669 <code>fast</code> LACP heartbeats are requested at a rate of once
670 per second causing connectivity problems to be detected more
671 quickly. In <code>slow</code> mode, heartbeats are requested at
672 a rate of once every 30 seconds.</p>
674 <p>Users may manually set a heartbeat transmission rate to increase
675 the fault detection speed further. When manually set, OVS
676 expects the partner switch to be configured with the same
677 transmission rate. Manually setting <code>lacp-time</code> to
678 something other than <code>fast</code> or <code>slow</code> is
679 not supported by the LACP specification.</p>
681 <dt><code>lacp-strict</code></dt>
682 <dd> When <code>true</code>, configures this <ref table="Port"/> to
683 require successful LACP negotiations to enable any slaves.
684 Defaults to <code>false</code> which safely allows LACP to be used
685 with switches that do not support the protocol.</dd>
686 <dt><code>lacp-force-aggregatable</code></dt>
687 <dd> When <code>true</code>, forces all slaves managed by this
688 <ref table="Port"/> to advertise themselves as aggregatable even if
689 they normally wouldn't. Defaults to <code>false</code>.</dd>
695 <table name="Interface" title="One physical network device in a Port.">
696 An interface within a <ref table="Port"/>.
698 <group title="Core Features">
700 Interface name. Should be alphanumeric and no more than about 8 bytes
701 long. May be the same as the port name, for non-bonded ports. Must
702 otherwise be unique among the names of ports, interfaces, and bridges
707 <p>Ethernet address to set for this interface. If unset then the
708 default MAC address is used:</p>
710 <li>For the local interface, the default is the lowest-numbered MAC
711 address among the other bridge ports, either the value of the
712 <ref table="Port" column="mac"/> in its <ref table="Port"/> record,
713 if set, or its actual MAC (for bonded ports, the MAC of its slave
714 whose name is first in alphabetical order). Internal ports and
715 bridge ports that are used as port mirroring destinations (see the
716 <ref table="Mirror"/> table) are ignored.</li>
717 <li>For other internal interfaces, the default MAC is randomly
719 <li>External interfaces typically have a MAC address associated with
722 <p>Some interfaces may not have a software-controllable MAC
726 <column name="ofport">
727 <p>OpenFlow port number for this interface. Unlike most columns, this
728 column's value should be set only by Open vSwitch itself. Other
729 clients should set this column to an empty set (the default) when
730 creating an <ref table="Interface"/>.</p>
731 <p>Open vSwitch populates this column when the port number becomes
732 known. If the interface is successfully added,
733 <ref column="ofport"/> will be set to a number between 1 and 65535
734 (generally either in the range 1 to 65279, inclusive, or 65534, the
735 port number for the OpenFlow ``local port''). If the interface
736 cannot be added then Open vSwitch sets this column
741 <group title="System-Specific Details">
743 The interface type, one of:
745 <dt><code>system</code></dt>
746 <dd>An ordinary network device, e.g. <code>eth0</code> on Linux.
747 Sometimes referred to as ``external interfaces'' since they are
748 generally connected to hardware external to that on which the Open
749 vSwitch is running. The empty string is a synonym for
750 <code>system</code>.</dd>
751 <dt><code>internal</code></dt>
752 <dd>A simulated network device that sends and receives traffic. An
753 internal interface whose <ref column="name"/> is the same as its
754 bridge's <ref table="Open_vSwitch" column="name"/> is called the
755 ``local interface.'' It does not make sense to bond an internal
756 interface, so the terms ``port'' and ``interface'' are often used
757 imprecisely for internal interfaces.</dd>
758 <dt><code>tap</code></dt>
759 <dd>A TUN/TAP device managed by Open vSwitch.</dd>
760 <dt><code>gre</code></dt>
761 <dd>An Ethernet over RFC 2890 Generic Routing Encapsulation over IPv4
762 tunnel. Each tunnel must be uniquely identified by the
763 combination of <code>remote_ip</code>, <code>local_ip</code>, and
764 <code>in_key</code>. Note that if two ports are defined that are
765 the same except one has an optional identifier and the other does
766 not, the more specific one is matched first. <code>in_key</code>
767 is considered more specific than <code>local_ip</code> if a port
768 defines one and another port defines the other. The following
769 options may be specified in the <ref column="options"/> column:
771 <dt><code>remote_ip</code></dt>
772 <dd>Required. The tunnel endpoint.</dd>
775 <dt><code>local_ip</code></dt>
776 <dd>Optional. The destination IP that received packets must
777 match. Default is to match all addresses.</dd>
780 <dt><code>in_key</code></dt>
781 <dd>Optional. The GRE key that received packets must contain.
782 It may either be a 32-bit number (no key and a key of 0 are
783 treated as equivalent) or the word <code>flow</code>. If
784 <code>flow</code> is specified then any key will be accepted
785 and the key will be placed in the <code>tun_id</code> field
786 for matching in the flow table. The ovs-ofctl manual page
787 contains additional information about matching fields in
788 OpenFlow flows. Default is no key.</dd>
791 <dt><code>out_key</code></dt>
792 <dd>Optional. The GRE key to be set on outgoing packets. It may
793 either be a 32-bit number or the word <code>flow</code>. If
794 <code>flow</code> is specified then the key may be set using
795 the <code>set_tunnel</code> Nicira OpenFlow vendor extension (0
796 is used in the absence of an action). The ovs-ofctl manual
797 page contains additional information about the Nicira OpenFlow
798 vendor extensions. Default is no key.</dd>
801 <dt><code>key</code></dt>
802 <dd>Optional. Shorthand to set <code>in_key</code> and
803 <code>out_key</code> at the same time.</dd>
806 <dt><code>tos</code></dt>
807 <dd>Optional. The value of the ToS bits to be set on the
808 encapsulating packet. It may also be the word
809 <code>inherit</code>, in which case the ToS will be copied from
810 the inner packet if it is IPv4 or IPv6 (otherwise it will be
811 0). Note that the ECN fields are always inherited. Default is
815 <dt><code>ttl</code></dt>
816 <dd>Optional. The TTL to be set on the encapsulating packet.
817 It may also be the word <code>inherit</code>, in which case the
818 TTL will be copied from the inner packet if it is IPv4 or IPv6
819 (otherwise it will be the system default, typically 64).
820 Default is the system default TTL.</dd>
823 <dt><code>csum</code></dt>
824 <dd>Optional. Compute GRE checksums on outgoing packets.
825 Checksums present on incoming packets will be validated
826 regardless of this setting. Note that GRE checksums
827 impose a significant performance penalty as they cover the
828 entire packet. As the contents of the packet is typically
829 covered by L3 and L4 checksums, this additional checksum only
830 adds value for the GRE and encapsulated Ethernet headers.
831 Default is disabled, set to <code>true</code> to enable.</dd>
834 <dt><code>df_inherit</code></dt>
835 <dd>Optional. If enabled, the Don't Fragment bit will be copied
836 from the inner IP headers (those of the encapsulated traffic)
837 to the outer (tunnel) headers. Default is disabled; set to
838 <code>true</code> to enable.</dd>
841 <dt><code>df_default</code></dt>
842 <dd>Optional. If enabled, the Don't Fragment bit will be set by
843 default on tunnel headers if the <code>df_inherit</code> option
844 is not set, or if the encapsulated packet is not IP. Default
845 is enabled; set to <code>false</code> to disable.</dd>
848 <dt><code>pmtud</code></dt>
849 <dd>Optional. Enable tunnel path MTU discovery. If enabled
850 ``ICMP Destination Unreachable - Fragmentation Needed''
851 messages will be generated for IPv4 packets with the DF bit set
852 and IPv6 packets above the minimum MTU if the packet size
853 exceeds the path MTU minus the size of the tunnel headers.
854 Note that this option causes behavior that is typically
855 reserved for routers and therefore is not entirely in
856 compliance with the IEEE 802.1D specification for bridges.
857 Default is enabled; set to <code>false</code> to disable.</dd>
860 <dt><code>header_cache</code></dt>
861 <dd>Optional. Enable caching of tunnel headers and the output
862 path. This can lead to a significant performance increase
863 without changing behavior. In general it should not be
864 necessary to adjust this setting. However, the caching can
865 bypass certain components of the IP stack (such as IP tables)
866 and it may be useful to disable it if these features are
867 required or as a debugging measure. Default is enabled, set to
868 <code>false</code> to disable.</dd>
871 <dt><code>ipsec_gre</code></dt>
872 <dd>An Ethernet over RFC 2890 Generic Routing Encapsulation
873 over IPv4 IPsec tunnel. Each tunnel (including those of type
874 <code>gre</code>) must be uniquely identified by the
875 combination of <code>remote_ip</code> and
876 <code>local_ip</code>. Note that if two ports are defined
877 that are the same except one has an optional identifier and
878 the other does not, the more specific one is matched first.
879 An authentication method of <code>peer_cert</code> or
880 <code>psk</code> must be defined. The following options may
881 be specified in the <ref column="options"/> column:
883 <dt><code>remote_ip</code></dt>
884 <dd>Required. The tunnel endpoint.</dd>
887 <dt><code>local_ip</code></dt>
888 <dd>Optional. The destination IP that received packets must
889 match. Default is to match all addresses.</dd>
892 <dt><code>peer_cert</code></dt>
893 <dd>Required for certificate authentication. A string
894 containing the peer's certificate in PEM format.
895 Additionally the host's certificate must be specified
896 with the <code>certificate</code> option.</dd>
899 <dt><code>certificate</code></dt>
900 <dd>Required for certificate authentication. The name of a
901 PEM file containing a certificate that will be presented
902 to the peer during authentication.</dd>
905 <dt><code>private_key</code></dt>
906 <dd>Optional for certificate authentication. The name of
907 a PEM file containing the private key associated with
908 <code>certificate</code>. If <code>certificate</code>
909 contains the private key, this option may be omitted.</dd>
912 <dt><code>psk</code></dt>
913 <dd>Required for pre-shared key authentication. Specifies a
914 pre-shared key for authentication that must be identical on
915 both sides of the tunnel.</dd>
918 <dt><code>in_key</code></dt>
919 <dd>Optional. The GRE key that received packets must contain.
920 It may either be a 32-bit number (no key and a key of 0 are
921 treated as equivalent) or the word <code>flow</code>. If
922 <code>flow</code> is specified then any key will be accepted
923 and the key will be placed in the <code>tun_id</code> field
924 for matching in the flow table. The ovs-ofctl manual page
925 contains additional information about matching fields in
926 OpenFlow flows. Default is no key.</dd>
929 <dt><code>out_key</code></dt>
930 <dd>Optional. The GRE key to be set on outgoing packets. It may
931 either be a 32-bit number or the word <code>flow</code>. If
932 <code>flow</code> is specified then the key may be set using
933 the <code>set_tunnel</code> Nicira OpenFlow vendor extension (0
934 is used in the absence of an action). The ovs-ofctl manual
935 page contains additional information about the Nicira OpenFlow
936 vendor extensions. Default is no key.</dd>
939 <dt><code>key</code></dt>
940 <dd>Optional. Shorthand to set <code>in_key</code> and
941 <code>out_key</code> at the same time.</dd>
944 <dt><code>tos</code></dt>
945 <dd>Optional. The value of the ToS bits to be set on the
946 encapsulating packet. It may also be the word
947 <code>inherit</code>, in which case the ToS will be copied from
948 the inner packet if it is IPv4 or IPv6 (otherwise it will be
949 0). Note that the ECN fields are always inherited. Default is
953 <dt><code>ttl</code></dt>
954 <dd>Optional. The TTL to be set on the encapsulating packet.
955 It may also be the word <code>inherit</code>, in which case the
956 TTL will be copied from the inner packet if it is IPv4 or IPv6
957 (otherwise it will be the system default, typically 64).
958 Default is the system default TTL.</dd>
961 <dt><code>csum</code></dt>
962 <dd>Optional. Compute GRE checksums on outgoing packets.
963 Checksums present on incoming packets will be validated
964 regardless of this setting. Note that GRE checksums
965 impose a significant performance penalty as they cover the
966 entire packet. As the contents of the packet is typically
967 covered by L3 and L4 checksums, this additional checksum only
968 adds value for the GRE and encapsulated Ethernet headers.
969 Default is disabled, set to <code>true</code> to enable.</dd>
972 <dt><code>df_inherit</code></dt>
973 <dd>Optional. If enabled, the Don't Fragment bit will be copied
974 from the inner IP headers (those of the encapsulated traffic)
975 to the outer (tunnel) headers. Default is disabled; set to
976 <code>true</code> to enable.</dd>
979 <dt><code>df_default</code></dt>
980 <dd>Optional. If enabled, the Don't Fragment bit will be set by
981 default on tunnel headers if the <code>df_inherit</code> option
982 is not set, or if the encapsulated packet is not IP. Default
983 is enabled; set to <code>false</code> to disable.</dd>
986 <dt><code>pmtud</code></dt>
987 <dd>Optional. Enable tunnel path MTU discovery. If enabled
988 ``ICMP Destination Unreachable - Fragmentation Needed''
989 messages will be generated for IPv4 packets with the DF bit set
990 and IPv6 packets above the minimum MTU if the packet size
991 exceeds the path MTU minus the size of the tunnel headers.
992 Note that this option causes behavior that is typically
993 reserved for routers and therefore is not entirely in
994 compliance with the IEEE 802.1D specification for bridges.
995 Default is enabled; set to <code>false</code> to disable.</dd>
998 <dt><code>capwap</code></dt>
999 <dd>Ethernet tunneling over the UDP transport portion of CAPWAP
1000 (RFC 5415). This allows interoperability with certain switches
1001 where GRE is not available. Note that only the tunneling component
1002 of the protocol is implemented. Due to the non-standard use of
1003 CAPWAP, UDP ports 58881 and 58882 are used as the source and
1004 destination ports respectively. Each tunnel must be uniquely
1005 identified by the combination of <code>remote_ip</code> and
1006 <code>local_ip</code>. If two ports are defined that are the same
1007 except one includes <code>local_ip</code> and the other does not,
1008 the more specific one is matched first. CAPWAP support is not
1009 available on all platforms. Currently it is only supported in the
1010 Linux kernel module with kernel versions >= 2.6.25. The following
1011 options may be specified in the <ref column="options"/> column:
1013 <dt><code>remote_ip</code></dt>
1014 <dd>Required. The tunnel endpoint.</dd>
1017 <dt><code>local_ip</code></dt>
1018 <dd>Optional. The destination IP that received packets must
1019 match. Default is to match all addresses.</dd>
1022 <dt><code>tos</code></dt>
1023 <dd>Optional. The value of the ToS bits to be set on the
1024 encapsulating packet. It may also be the word
1025 <code>inherit</code>, in which case the ToS will be copied from
1026 the inner packet if it is IPv4 or IPv6 (otherwise it will be
1027 0). Note that the ECN fields are always inherited. Default is
1031 <dt><code>ttl</code></dt>
1032 <dd>Optional. The TTL to be set on the encapsulating packet.
1033 It may also be the word <code>inherit</code>, in which case the
1034 TTL will be copied from the inner packet if it is IPv4 or IPv6
1035 (otherwise it will be the system default, typically 64).
1036 Default is the system default TTL.</dd>
1039 <dt><code>df_inherit</code></dt>
1040 <dd>Optional. If enabled, the Don't Fragment bit will be copied
1041 from the inner IP headers (those of the encapsulated traffic)
1042 to the outer (tunnel) headers. Default is disabled; set to
1043 <code>true</code> to enable.</dd>
1046 <dt><code>df_default</code></dt>
1047 <dd>Optional. If enabled, the Don't Fragment bit will be set by
1048 default on tunnel headers if the <code>df_inherit</code> option
1049 is not set, or if the encapsulated packet is not IP. Default
1050 is enabled; set to <code>false</code> to disable.</dd>
1053 <dt><code>pmtud</code></dt>
1054 <dd>Optional. Enable tunnel path MTU discovery. If enabled
1055 ``ICMP Destination Unreachable - Fragmentation Needed''
1056 messages will be generated for IPv4 packets with the DF bit set
1057 and IPv6 packets above the minimum MTU if the packet size
1058 exceeds the path MTU minus the size of the tunnel headers.
1059 Note that this option causes behavior that is typically
1060 reserved for routers and therefore is not entirely in
1061 compliance with the IEEE 802.1D specification for bridges.
1062 Default is enabled; set to <code>false</code> to disable.</dd>
1065 <dt><code>header_cache</code></dt>
1066 <dd>Optional. Enable caching of tunnel headers and the output
1067 path. This can lead to a significant performance increase
1068 without changing behavior. In general it should not be
1069 necessary to adjust this setting. However, the caching can
1070 bypass certain components of the IP stack (such as IP tables)
1071 and it may be useful to disable it if these features are
1072 required or as a debugging measure. Default is enabled, set to
1073 <code>false</code> to disable.</dd>
1076 <dt><code>patch</code></dt>
1079 A pair of virtual devices that act as a patch cable. The <ref
1080 column="options"/> column must have the following key-value pair:
1083 <dt><code>peer</code></dt>
1085 The <ref column="name"/> of the <ref table="Interface"/> for
1086 the other side of the patch. The named <ref
1087 table="Interface"/>'s own <code>peer</code> option must specify
1088 this <ref table="Interface"/>'s name. That is, the two patch
1089 interfaces must have reversed <ref column="name"/> and
1090 <code>peer</code> values.
1094 <dt><code>null</code></dt>
1095 <dd>An ignored interface.</dd>
1099 <column name="options">
1100 Configuration options whose interpretation varies based on
1101 <ref column="type"/>.
1105 <group title="Interface Status">
1107 Status information about interfaces attached to bridges, updated every
1108 5 seconds. Not all interfaces have all of these properties; virtual
1109 interfaces don't have a link speed, for example. Non-applicable
1110 columns will have empty values.
1112 <column name="admin_state">
1114 The administrative state of the physical network link.
1118 <column name="link_state">
1120 The observed state of the physical network link. This is ordinarily
1121 the link's carrier status. If the interface's <ref table="Port"/> is
1122 a bond configured for miimon monitoring, it is instead the network
1123 link's miimon status.
1127 <column name="link_speed">
1129 The negotiated speed of the physical network link.
1130 Valid values are positive integers greater than 0.
1134 <column name="duplex">
1136 The duplex mode of the physical network link.
1142 The MTU (maximum transmission unit); i.e. the largest
1143 amount of data that can fit into a single Ethernet frame.
1144 The standard Ethernet MTU is 1500 bytes. Some physical media
1145 and many kinds of virtual interfaces can be configured with
1149 This column will be empty for an interface that does not
1150 have an MTU as, for example, some kinds of tunnels do not.
1154 <column name="status">
1156 Key-value pairs that report port status. Supported status
1157 values are <code>type</code>-dependent; some interfaces may not have
1158 a valid <code>driver_name</code>, for example.
1160 <p>The currently defined key-value pairs are:</p>
1162 <dt><code>driver_name</code></dt>
1163 <dd>The name of the device driver controlling the network
1167 <dt><code>driver_version</code></dt>
1168 <dd>The version string of the device driver controlling the
1169 network adapter.</dd>
1172 <dt><code>firmware_version</code></dt>
1173 <dd>The version string of the network adapter's firmware, if
1177 <dt><code>source_ip</code></dt>
1178 <dd>The source IP address used for an IPv4 tunnel end-point,
1179 such as <code>gre</code> or <code>capwap</code>.</dd>
1182 <dt><code>tunnel_egress_iface</code></dt>
1183 <dd>Egress interface for tunnels. Currently only relevant for GRE
1184 and CAPWAP tunnels. On Linux systems, this column will show
1185 the name of the interface which is responsible for routing
1186 traffic destined for the configured <code>remote_ip</code>.
1187 This could be an internal interface such as a bridge port.</dd>
1190 <dt><code>tunnel_egress_iface_carrier</code></dt>
1191 <dd>Whether a carrier is detected on <ref
1192 column="tunnel_egress_iface"/>. Valid values are <code>down</code>
1193 and <code>up</code>.</dd>
1198 <group title="Ingress Policing">
1200 These settings control ingress policing for packets received on this
1201 interface. On a physical interface, this limits the rate at which
1202 traffic is allowed into the system from the outside; on a virtual
1203 interface (one connected to a virtual machine), this limits the rate at
1204 which the VM is able to transmit.
1207 Policing is a simple form of quality-of-service that simply drops
1208 packets received in excess of the configured rate. Due to its
1209 simplicity, policing is usually less accurate and less effective than
1210 egress QoS (which is configured using the <ref table="QoS"/> and <ref
1211 table="Queue"/> tables).
1214 Policing is currently implemented only on Linux. The Linux
1215 implementation uses a simple ``token bucket'' approach:
1219 The size of the bucket corresponds to <ref
1220 column="ingress_policing_burst"/>. Initially the bucket is full.
1223 Whenever a packet is received, its size (converted to tokens) is
1224 compared to the number of tokens currently in the bucket. If the
1225 required number of tokens are available, they are removed and the
1226 packet is forwarded. Otherwise, the packet is dropped.
1229 Whenever it is not full, the bucket is refilled with tokens at the
1230 rate specified by <ref column="ingress_policing_rate"/>.
1234 Policing interacts badly with some network protocols, and especially
1235 with fragmented IP packets. Suppose that there is enough network
1236 activity to keep the bucket nearly empty all the time. Then this token
1237 bucket algorithm will forward a single packet every so often, with the
1238 period depending on packet size and on the configured rate. All of the
1239 fragments of an IP packets are normally transmitted back-to-back, as a
1240 group. In such a situation, therefore, only one of these fragments
1241 will be forwarded and the rest will be dropped. IP does not provide
1242 any way for the intended recipient to ask for only the remaining
1243 fragments. In such a case there are two likely possibilities for what
1244 will happen next: either all of the fragments will eventually be
1245 retransmitted (as TCP will do), in which case the same problem will
1246 recur, or the sender will not realize that its packet has been dropped
1247 and data will simply be lost (as some UDP-based protocols will do).
1248 Either way, it is possible that no forward progress will ever occur.
1250 <column name="ingress_policing_rate">
1252 Maximum rate for data received on this interface, in kbps. Data
1253 received faster than this rate is dropped. Set to <code>0</code>
1254 (the default) to disable policing.
1258 <column name="ingress_policing_burst">
1259 <p>Maximum burst size for data received on this interface, in kb. The
1260 default burst size if set to <code>0</code> is 1000 kb. This value
1261 has no effect if <ref column="ingress_policing_rate"/>
1262 is <code>0</code>.</p>
1264 Specifying a larger burst size lets the algorithm be more forgiving,
1265 which is important for protocols like TCP that react severely to
1266 dropped packets. The burst size should be at least the size of the
1267 interface's MTU. Specifying a value that is numerically at least as
1268 large as 10% of <ref column="ingress_policing_rate"/> helps TCP come
1269 closer to achieving the full rate.
1274 <group title="Other Features">
1276 <column name="monitor">
1277 Connectivity monitor configuration for this interface.
1280 <column name="lacp_current">
1281 Boolean value indicating LACP status for this interface. If true, this
1282 interface has current LACP information about its LACP partner. This
1283 information may be used to monitor the health of interfaces in a LACP
1284 enabled port. This column will be empty if LACP is not enabled.
1287 <column name="external_ids">
1288 Key-value pairs for use by external frameworks that integrate
1289 with Open vSwitch, rather than by Open vSwitch itself. System
1290 integrators should either use the Open vSwitch development
1291 mailing list to coordinate on common key-value definitions, or
1292 choose key names that are likely to be unique. The currently
1293 defined common key-value pairs are:
1295 <dt><code>attached-mac</code></dt>
1297 The MAC address programmed into the ``virtual hardware'' for this
1298 interface, in the form
1299 <var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>:<var>xx</var>.
1300 For Citrix XenServer, this is the value of the <code>MAC</code>
1301 field in the VIF record for this interface.</dd>
1302 <dt><code>iface-id</code></dt>
1303 <dd>A system-unique identifier for the interface. On XenServer,
1304 this will commonly be the same as <code>xs-vif-uuid</code>.</dd>
1307 Additionally the following key-value pairs specifically
1308 apply to an interface that represents a virtual Ethernet interface
1309 connected to a virtual machine. These key-value pairs should not be
1310 present for other types of interfaces. Keys whose names end
1311 in <code>-uuid</code> have values that uniquely identify the entity
1312 in question. For a Citrix XenServer hypervisor, these values are
1313 UUIDs in RFC 4122 format. Other hypervisors may use other
1316 <p>The currently defined key-value pairs for XenServer are:</p>
1318 <dt><code>xs-vif-uuid</code></dt>
1319 <dd>The virtual interface associated with this interface.</dd>
1320 <dt><code>xs-network-uuid</code></dt>
1321 <dd>The virtual network to which this interface is attached.</dd>
1322 <dt><code>xs-vm-uuid</code></dt>
1323 <dd>The VM to which this interface belongs.</dd>
1327 <column name="other_config">
1328 Key-value pairs for rarely used interface features.
1330 <dt><code>bond-stable-id</code></dt>
1331 <dd> A positive integer using in <code>stable</code> bond mode to
1332 make slave selection decisions. Allocating
1333 <code>bond-stable-id</code>s consistently across interfaces
1334 participating in a bond will guarantee consistent slave selection
1335 decisions across ovs-vswitchd instances when using
1336 <code>stable</code> bonding mode.</dd>
1337 <dt><code>lacp-port-id</code></dt>
1338 <dd> The LACP port ID of this <ref table="Interface"/>. Port IDs are
1339 used in LACP negotiations to identify individual ports
1340 participating in a bond. Must be a number between 1 and
1342 <dt><code>lacp-port-priority</code></dt>
1343 <dd> The LACP port priority of this <ref table="Interface"/>. In
1344 LACP negotiations <ref table="Interface"/>s with numerically lower
1345 priorities are preferred for aggregation. Must be a number between
1347 <dt><code>lacp-aggregation-key</code></dt>
1348 <dd> The LACP aggregation key of this <ref table="Interface"/>.
1349 <ref table="Interface"/>s with different aggregation keys may not
1350 be active within a given <ref table="Port"/> at the same time. Must
1351 be a number between 1 and 65535.</dd>
1355 <column name="statistics">
1357 Key-value pairs that report interface statistics. The current
1358 implementation updates these counters periodically. In the future,
1359 we plan to, instead, update them when an interface is created, when
1360 they are queried (e.g. using an OVSDB <code>select</code> operation),
1361 and just before an interface is deleted due to virtual interface
1362 hot-unplug or VM shutdown, and perhaps at other times, but not on any
1363 regular periodic basis.</p>
1365 The currently defined key-value pairs are listed below. These are
1366 the same statistics reported by OpenFlow in its <code>struct
1367 ofp_port_stats</code> structure. If an interface does not support a
1368 given statistic, then that pair is omitted.</p>
1371 Successful transmit and receive counters:
1373 <dt><code>rx_packets</code></dt>
1374 <dd>Number of received packets.</dd>
1375 <dt><code>rx_bytes</code></dt>
1376 <dd>Number of received bytes.</dd>
1377 <dt><code>tx_packets</code></dt>
1378 <dd>Number of transmitted packets.</dd>
1379 <dt><code>tx_bytes</code></dt>
1380 <dd>Number of transmitted bytes.</dd>
1386 <dt><code>rx_dropped</code></dt>
1387 <dd>Number of packets dropped by RX.</dd>
1388 <dt><code>rx_frame_err</code></dt>
1389 <dd>Number of frame alignment errors.</dd>
1390 <dt><code>rx_over_err</code></dt>
1391 <dd>Number of packets with RX overrun.</dd>
1392 <dt><code>rx_crc_err</code></dt>
1393 <dd>Number of CRC errors.</dd>
1394 <dt><code>rx_errors</code></dt>
1396 Total number of receive errors, greater than or equal
1397 to the sum of the above.
1404 <dt><code>tx_dropped</code></dt>
1405 <dd>Number of packets dropped by TX.</dd>
1406 <dt><code>collisions</code></dt>
1407 <dd>Number of collisions.</dd>
1408 <dt><code>tx_errors</code></dt>
1410 Total number of transmit errors, greater
1411 than or equal to the sum of the above.
1420 <table name="QoS" title="Quality of Service configuration">
1421 <p>Quality of Service (QoS) configuration for each Port that
1424 <column name="type">
1425 <p>The type of QoS to implement. The <ref table="Open_vSwitch"
1426 column="capabilities"/> column in the <ref table="Open_vSwitch"/> table
1427 identifies the types that a switch actually supports. The currently
1428 defined types are listed below:</p>
1430 <dt><code>linux-htb</code></dt>
1432 Linux ``hierarchy token bucket'' classifier. See tc-htb(8) (also at
1433 <code>http://linux.die.net/man/8/tc-htb</code>) and the HTB manual
1434 (<code>http://luxik.cdi.cz/~devik/qos/htb/manual/userg.htm</code>)
1435 for information on how this classifier works and how to configure it.
1439 <dt><code>linux-hfsc</code></dt>
1441 Linux "Hierarchical Fair Service Curve" classifier.
1442 See <code>http://linux-ip.net/articles/hfsc.en/</code> for
1443 information on how this classifier works.
1448 <column name="queues">
1449 <p>A map from queue numbers to <ref table="Queue"/> records. The
1450 supported range of queue numbers depend on <ref column="type"/>. The
1451 queue numbers are the same as the <code>queue_id</code> used in
1452 OpenFlow in <code>struct ofp_action_enqueue</code> and other
1453 structures. Queue 0 is used by OpenFlow output actions that do not
1454 specify a specific queue.</p>
1457 <column name="other_config">
1458 <p>Key-value pairs for configuring QoS features that depend on
1459 <ref column="type"/>.</p>
1460 <p>The <code>linux-htb</code> and <code>linux-hfsc</code> classes support
1461 the following key-value pairs:</p>
1463 <dt><code>max-rate</code></dt>
1464 <dd>Maximum rate shared by all queued traffic, in bit/s.
1465 Optional. If not specified, for physical interfaces, the
1466 default is the link rate. For other interfaces or if the
1467 link rate cannot be determined, the default is currently 100
1472 <column name="external_ids">
1473 Key-value pairs for use by external frameworks that integrate with Open
1474 vSwitch, rather than by Open vSwitch itself. System integrators should
1475 either use the Open vSwitch development mailing list to coordinate on
1476 common key-value definitions, or choose key names that are likely to be
1477 unique. No common key-value pairs are currently defined.
1481 <table name="Queue" title="QoS output queue.">
1482 <p>A configuration for a port output queue, used in configuring Quality of
1483 Service (QoS) features. May be referenced by <ref column="queues"
1484 table="QoS"/> column in <ref table="QoS"/> table.</p>
1486 <column name="other_config">
1487 <p>Key-value pairs for configuring the output queue. The supported
1488 key-value pairs and their meanings depend on the <ref column="type"/>
1489 of the <ref column="QoS"/> records that reference this row.</p>
1490 <p>The key-value pairs defined for <ref table="QoS"/> <ref table="QoS"
1491 column="type"/> of <code>min-rate</code> are:</p>
1493 <dt><code>min-rate</code></dt>
1494 <dd>Minimum guaranteed bandwidth, in bit/s. Required. The
1495 floor value is 1500 bytes/s (12,000 bit/s).</dd>
1497 <p>The key-value pairs defined for <ref table="QoS"/> <ref table="QoS"
1498 column="type"/> of <code>linux-htb</code> are:</p>
1500 <dt><code>min-rate</code></dt>
1501 <dd>Minimum guaranteed bandwidth, in bit/s.</dd>
1502 <dt><code>max-rate</code></dt>
1503 <dd>Maximum allowed bandwidth, in bit/s. Optional. If specified, the
1504 queue's rate will not be allowed to exceed the specified value, even
1505 if excess bandwidth is available. If unspecified, defaults to no
1507 <dt><code>burst</code></dt>
1508 <dd>Burst size, in bits. This is the maximum amount of ``credits''
1509 that a queue can accumulate while it is idle. Optional. Details of
1510 the <code>linux-htb</code> implementation require a minimum burst
1511 size, so a too-small <code>burst</code> will be silently
1513 <dt><code>priority</code></dt>
1514 <dd>A nonnegative 32-bit integer. Defaults to 0 if
1515 unspecified. A queue with a smaller <code>priority</code>
1516 will receive all the excess bandwidth that it can use before
1517 a queue with a larger value receives any. Specific priority
1518 values are unimportant; only relative ordering matters.</dd>
1520 <p>The key-value pairs defined for <ref table="QoS"/> <ref table="QoS"
1521 column="type"/> of <code>linux-hfsc</code> are:</p>
1523 <dt><code>min-rate</code></dt>
1524 <dd>Minimum guaranteed bandwidth, in bit/s.</dd>
1525 <dt><code>max-rate</code></dt>
1526 <dd>Maximum allowed bandwidth, in bit/s. Optional. If specified, the
1527 queue's rate will not be allowed to exceed the specified value, even
1528 if excess bandwidth is available. If unspecified, defaults to no
1533 <column name="external_ids">
1534 Key-value pairs for use by external frameworks that integrate with Open
1535 vSwitch, rather than by Open vSwitch itself. System integrators should
1536 either use the Open vSwitch development mailing list to coordinate on
1537 common key-value definitions, or choose key names that are likely to be
1538 unique. No common key-value pairs are currently defined.
1542 <table name="Monitor" title="Connectivity Monitor configuration">
1544 A <ref table="Monitor"/> attaches to an <ref table="Interface"/> to
1545 implement 802.1ag Connectivity Fault Management (CFM). CFM allows a
1546 group of Maintenance Points (MPs) called a Maintenance Association (MA)
1547 to detect connectivity problems with each other. MPs within a MA should
1548 have complete and exclusive interconnectivity. This is verified by
1549 occasionally broadcasting Continuity Check Messages (CCMs) at a
1550 configurable transmission interval. A <ref table="Monitor"/> is
1551 responsible for collecting data about other MPs in its MA and
1555 <group title="Monitor Configuration">
1556 <column name="mpid">
1557 A Maintenance Point ID (MPID) uniquely identifies each endpoint within
1558 a Maintenance Association (see <ref column="ma_name"/>). The MPID is
1559 used to identify this <ref table="Monitor"/> to other endpoints in the
1563 <column name="remote_mps">
1564 A set of <ref table="Maintenance_Points"/> which this
1565 <ref table="Monitor"/> should have connectivity to. If this
1566 <ref table="Monitor"/> does not have connectivity to any MPs in this
1567 set, or has connectivity to any MPs not in this set, a fault is
1571 <column name="ma_name">
1572 A Maintenance Association (MA) name pairs with a Maintenance Domain
1573 (MD) name to uniquely identify a MA. A MA is a group of endpoints who
1574 have complete and exclusive interconnectivity. Defaults to
1575 <code>ovs</code> if unset.
1578 <column name="md_name">
1579 A Maintenance Domain name pairs with a Maintenance Association name to
1580 uniquely identify a MA. Defaults to <code>ovs</code> if unset.
1583 <column name="interval">
1584 The transmission interval of CCMs in milliseconds. Three missed CCMs
1585 indicate a connectivity fault. Defaults to 1000ms.
1589 <group title="Monitor Status">
1590 <column name="fault">
1591 Indicates a Connectivity Fault caused by a configuration error, a down
1592 remote MP, or unexpected connectivity to a remote MAID or remote MP.
1597 <table name="Maintenance_Point" title="Maintenance Point configuration">
1599 A <ref table="Maintenance_Point"/> represents a MP which a
1600 <ref table="Monitor"/> has or should have connectivity to.
1603 <group title="Maintenance_Point Configuration">
1604 <column name="mpid">
1605 A Maintenance Point ID (MPID) uniquely identifies each endpoint within
1606 a Maintenance Association. All MPs within a MA should have a unique
1611 <group title="Maintenance_Point Status">
1612 <column name="fault">
1613 Indicates a connectivity fault.
1618 <table name="Mirror" title="Port mirroring (SPAN/RSPAN).">
1619 <p>A port mirror within a <ref table="Bridge"/>.</p>
1620 <p>A port mirror configures a bridge to send selected frames to special
1621 ``mirrored'' ports, in addition to their normal destinations. Mirroring
1622 traffic may also be referred to as SPAN or RSPAN, depending on the
1623 mechanism used for delivery.</p>
1625 <column name="name">
1626 Arbitrary identifier for the <ref table="Mirror"/>.
1629 <group title="Selecting Packets for Mirroring">
1631 To be selected for mirroring, a given packet must enter or leave the
1632 bridge through a selected port and it must also be in one of the
1636 <column name="select_all">
1637 If true, every packet arriving or departing on any port is
1638 selected for mirroring.
1641 <column name="select_dst_port">
1642 Ports on which departing packets are selected for mirroring.
1645 <column name="select_src_port">
1646 Ports on which arriving packets are selected for mirroring.
1649 <column name="select_vlan">
1650 VLANs on which packets are selected for mirroring. An empty set
1651 selects packets on all VLANs.
1655 <group title="Mirroring Destination Configuration">
1657 These columns are mutually exclusive. Exactly one of them must be
1661 <column name="output_port">
1662 <p>Output port for selected packets, if nonempty.</p>
1663 <p>Specifying a port for mirror output reserves that port exclusively
1664 for mirroring. No frames other than those selected for mirroring
1665 will be forwarded to the port, and any frames received on the port
1666 will be discarded.</p>
1667 <p>This type of mirroring is sometimes called SPAN.</p>
1670 <column name="output_vlan">
1671 <p>Output VLAN for selected packets, if nonempty.</p>
1672 <p>The frames will be sent out all ports that trunk
1673 <ref column="output_vlan"/>, as well as any ports with implicit VLAN
1674 <ref column="output_vlan"/>. When a mirrored frame is sent out a
1675 trunk port, the frame's VLAN tag will be set to
1676 <ref column="output_vlan"/>, replacing any existing tag; when it is
1677 sent out an implicit VLAN port, the frame will not be tagged. This
1678 type of mirroring is sometimes called RSPAN.</p>
1679 <p><em>Please note:</em> Mirroring to a VLAN can disrupt a network that
1680 contains unmanaged switches. Consider an unmanaged physical switch
1681 with two ports: port 1, connected to an end host, and port 2,
1682 connected to an Open vSwitch configured to mirror received packets
1683 into VLAN 123 on port 2. Suppose that the end host sends a packet on
1684 port 1 that the physical switch forwards to port 2. The Open vSwitch
1685 forwards this packet to its destination and then reflects it back on
1686 port 2 in VLAN 123. This reflected packet causes the unmanaged
1687 physical switch to replace the MAC learning table entry, which
1688 correctly pointed to port 1, with one that incorrectly points to port
1689 2. Afterward, the physical switch will direct packets destined for
1690 the end host to the Open vSwitch on port 2, instead of to the end
1691 host on port 1, disrupting connectivity. If mirroring to a VLAN is
1692 desired in this scenario, then the physical switch must be replaced
1693 by one that learns Ethernet addresses on a per-VLAN basis. In
1694 addition, learning should be disabled on the VLAN containing mirrored
1695 traffic. If this is not done then intermediate switches will learn
1696 the MAC address of each end host from the mirrored traffic. If
1697 packets being sent to that end host are also mirrored, then they will
1698 be dropped since the switch will attempt to send them out the input
1699 port. Disabling learning for the VLAN will cause the switch to
1700 correctly send the packet out all ports configured for that VLAN. If
1701 Open vSwitch is being used as an intermediate switch, learning can be
1702 disabled by adding the mirrored VLAN to <ref column="flood_vlans"/>
1703 in the appropriate <ref table="Bridge"/> table or tables.</p>
1707 <group title="Other Features">
1708 <column name="external_ids">
1709 Key-value pairs for use by external frameworks that integrate with Open
1710 vSwitch, rather than by Open vSwitch itself. System integrators should
1711 either use the Open vSwitch development mailing list to coordinate on
1712 common key-value definitions, or choose key names that are likely to be
1713 unique. No common key-value pairs are currently defined.
1718 <table name="Controller" title="OpenFlow controller configuration.">
1719 <p>An OpenFlow controller.</p>
1722 Open vSwitch supports two kinds of OpenFlow controllers:
1726 <dt>Primary controllers</dt>
1729 This is the kind of controller envisioned by the OpenFlow 1.0
1730 specification. Usually, a primary controller implements a network
1731 policy by taking charge of the switch's flow table.
1735 Open vSwitch initiates and maintains persistent connections to
1736 primary controllers, retrying the connection each time it fails or
1737 drops. The <ref table="Bridge" column="fail_mode"/> column in the
1738 <ref table="Bridge"/> table applies to primary controllers.
1742 Open vSwitch permits a bridge to have any number of primary
1743 controllers. When multiple controllers are configured, Open
1744 vSwitch connects to all of them simultaneously. Because
1745 OpenFlow 1.0 does not specify how multiple controllers
1746 coordinate in interacting with a single switch, more than
1747 one primary controller should be specified only if the
1748 controllers are themselves designed to coordinate with each
1749 other. (The Nicira-defined <code>NXT_ROLE</code> OpenFlow
1750 vendor extension may be useful for this.)
1753 <dt>Service controllers</dt>
1756 These kinds of OpenFlow controller connections are intended for
1757 occasional support and maintenance use, e.g. with
1758 <code>ovs-ofctl</code>. Usually a service controller connects only
1759 briefly to inspect or modify some of a switch's state.
1763 Open vSwitch listens for incoming connections from service
1764 controllers. The service controllers initiate and, if necessary,
1765 maintain the connections from their end. The <ref table="Bridge"
1766 column="fail_mode"/> column in the <ref table="Bridge"/> table does
1767 not apply to service controllers.
1771 Open vSwitch supports configuring any number of service controllers.
1777 The <ref column="target"/> determines the type of controller.
1780 <group title="Core Features">
1781 <column name="target">
1782 <p>Connection method for controller.</p>
1784 The following connection methods are currently supported for primary
1788 <dt><code>ssl:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
1790 <p>The specified SSL <var>port</var> (default: 6633) on the host at
1791 the given <var>ip</var>, which must be expressed as an IP address
1792 (not a DNS name). The <ref table="Open_vSwitch" column="ssl"/>
1793 column in the <ref table="Open_vSwitch"/> table must point to a
1794 valid SSL configuration when this form is used.</p>
1795 <p>SSL support is an optional feature that is not always built as
1796 part of Open vSwitch.</p>
1798 <dt><code>tcp:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
1799 <dd>The specified TCP <var>port</var> (default: 6633) on the host at
1800 the given <var>ip</var>, which must be expressed as an IP address
1801 (not a DNS name).</dd>
1804 The following connection methods are currently supported for service
1808 <dt><code>pssl:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
1811 Listens for SSL connections on the specified TCP <var>port</var>
1812 (default: 6633). If <var>ip</var>, which must be expressed as an
1813 IP address (not a DNS name), is specified, then connections are
1814 restricted to the specified local IP address.
1817 The <ref table="Open_vSwitch" column="ssl"/> column in the <ref
1818 table="Open_vSwitch"/> table must point to a valid SSL
1819 configuration when this form is used.
1821 <p>SSL support is an optional feature that is not always built as
1822 part of Open vSwitch.</p>
1824 <dt><code>ptcp:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
1826 Listens for connections on the specified TCP <var>port</var>
1827 (default: 6633). If <var>ip</var>, which must be expressed as an
1828 IP address (not a DNS name), is specified, then connections are
1829 restricted to the specified local IP address.
1832 <p>When multiple controllers are configured for a single bridge, the
1833 <ref column="target"/> values must be unique. Duplicate
1834 <ref column="target"/> values yield unspecified results.</p>
1837 <column name="connection_mode">
1838 <p>If it is specified, this setting must be one of the following
1839 strings that describes how Open vSwitch contacts this OpenFlow
1840 controller over the network:</p>
1843 <dt><code>in-band</code></dt>
1844 <dd>In this mode, this controller's OpenFlow traffic travels over the
1845 bridge associated with the controller. With this setting, Open
1846 vSwitch allows traffic to and from the controller regardless of the
1847 contents of the OpenFlow flow table. (Otherwise, Open vSwitch
1848 would never be able to connect to the controller, because it did
1849 not have a flow to enable it.) This is the most common connection
1850 mode because it is not necessary to maintain two independent
1852 <dt><code>out-of-band</code></dt>
1853 <dd>In this mode, OpenFlow traffic uses a control network separate
1854 from the bridge associated with this controller, that is, the
1855 bridge does not use any of its own network devices to communicate
1856 with the controller. The control network must be configured
1857 separately, before or after <code>ovs-vswitchd</code> is started.
1861 <p>If not specified, the default is implementation-specific.</p>
1865 <group title="Controller Failure Detection and Handling">
1866 <column name="max_backoff">
1867 Maximum number of milliseconds to wait between connection attempts.
1868 Default is implementation-specific.
1871 <column name="inactivity_probe">
1872 Maximum number of milliseconds of idle time on connection to
1873 controller before sending an inactivity probe message. If Open
1874 vSwitch does not communicate with the controller for the specified
1875 number of seconds, it will send a probe. If a response is not
1876 received for the same additional amount of time, Open vSwitch
1877 assumes the connection has been broken and attempts to reconnect.
1878 Default is implementation-specific. A value of 0 disables
1883 <group title="OpenFlow Rate Limiting">
1884 <column name="controller_rate_limit">
1885 <p>The maximum rate at which packets in unknown flows will be
1886 forwarded to the OpenFlow controller, in packets per second. This
1887 feature prevents a single bridge from overwhelming the controller.
1888 If not specified, the default is implementation-specific.</p>
1889 <p>In addition, when a high rate triggers rate-limiting, Open
1890 vSwitch queues controller packets for each port and transmits
1891 them to the controller at the configured rate. The number of
1892 queued packets is limited by
1893 the <ref column="controller_burst_limit"/> value. The packet
1894 queue is shared fairly among the ports on a bridge.</p><p>Open
1895 vSwitch maintains two such packet rate-limiters per bridge.
1896 One of these applies to packets sent up to the controller
1897 because they do not correspond to any flow. The other applies
1898 to packets sent up to the controller by request through flow
1899 actions. When both rate-limiters are filled with packets, the
1900 actual rate that packets are sent to the controller is up to
1901 twice the specified rate.</p>
1904 <column name="controller_burst_limit">
1905 In conjunction with <ref column="controller_rate_limit"/>,
1906 the maximum number of unused packet credits that the bridge will
1907 allow to accumulate, in packets. If not specified, the default
1908 is implementation-specific.
1912 <group title="Additional In-Band Configuration">
1913 <p>These values are considered only in in-band control mode (see
1914 <ref column="connection_mode"/>).</p>
1916 <p>When multiple controllers are configured on a single bridge, there
1917 should be only one set of unique values in these columns. If different
1918 values are set for these columns in different controllers, the effect
1921 <column name="local_ip">
1922 The IP address to configure on the local port,
1923 e.g. <code>192.168.0.123</code>. If this value is unset, then
1924 <ref column="local_netmask"/> and <ref column="local_gateway"/> are
1928 <column name="local_netmask">
1929 The IP netmask to configure on the local port,
1930 e.g. <code>255.255.255.0</code>. If <ref column="local_ip"/> is set
1931 but this value is unset, then the default is chosen based on whether
1932 the IP address is class A, B, or C.
1935 <column name="local_gateway">
1936 The IP address of the gateway to configure on the local port, as a
1937 string, e.g. <code>192.168.0.1</code>. Leave this column unset if
1938 this network has no gateway.
1942 <group title="Other Features">
1943 <column name="external_ids">
1944 Key-value pairs for use by external frameworks that integrate with Open
1945 vSwitch, rather than by Open vSwitch itself. System integrators should
1946 either use the Open vSwitch development mailing list to coordinate on
1947 common key-value definitions, or choose key names that are likely to be
1948 unique. No common key-value pairs are currently defined.
1952 <group title="Controller Status">
1953 <column name="is_connected">
1954 <code>true</code> if currently connected to this controller,
1955 <code>false</code> otherwise.
1958 <column name="role">
1959 <p>The level of authority this controller has on the associated
1960 bridge. Possible values are:</p>
1962 <dt><code>other</code></dt>
1963 <dd>Allows the controller access to all OpenFlow features.</dd>
1964 <dt><code>master</code></dt>
1965 <dd>Equivalent to <code>other</code>, except that there may be at
1966 most one master controller at a time. When a controller configures
1967 itself as <code>master</code>, any existing master is demoted to
1968 the <code>slave</code>role.</dd>
1969 <dt><code>slave</code></dt>
1970 <dd>Allows the controller read-only access to OpenFlow features.
1971 Attempts to modify the flow table will be rejected with an
1972 error. Slave controllers do not receive OFPT_PACKET_IN or
1973 OFPT_FLOW_REMOVED messages, but they do receive OFPT_PORT_STATUS
1978 <column name="status">
1979 <p>Key-value pairs that report controller status.</p>
1981 <dt><code>last_error</code></dt>
1982 <dd>A human-readable description of the last error on the connection
1983 to the controller; i.e. <code>strerror(errno)</code>. This key
1984 will exist only if an error has occurred.</dd>
1985 <dt><code>state</code></dt>
1986 <dd>The state of the connection to the controller. Possible values
1987 are: <code>VOID</code> (connection is disabled),
1988 <code>BACKOFF</code> (attempting to reconnect at an increasing
1989 period), <code>CONNECTING</code> (attempting to connect),
1990 <code>ACTIVE</code> (connected, remote host responsive), and
1991 <code>IDLE</code> (remote host idle, sending keep-alive). These
1992 values may change in the future. They are provided only for human
1994 <dt><code>sec_since_connect</code></dt>
1995 <dd>The amount of time since this controller last successfully
1996 connected to the switch (in seconds). Value is empty if controller
1997 has never successfully connected.</dd>
1998 <dt><code>sec_since_disconnect</code></dt>
1999 <dd>The amount of time since this controller last disconnected from
2000 the switch (in seconds). Value is empty if controller has never
2007 <table name="Manager" title="OVSDB management connection.">
2009 Configuration for a database connection to an Open vSwitch database
2014 This table primarily configures the Open vSwitch database
2015 (<code>ovsdb-server</code>), not the Open vSwitch switch
2016 (<code>ovs-vswitchd</code>). The switch does read the table to determine
2017 what connections should be treated as in-band.
2021 The Open vSwitch database server can initiate and maintain active
2022 connections to remote clients. It can also listen for database
2026 <group title="Core Features">
2027 <column name="target">
2028 <p>Connection method for managers.</p>
2030 The following connection methods are currently supported:
2033 <dt><code>ssl:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2036 The specified SSL <var>port</var> (default: 6632) on the host at
2037 the given <var>ip</var>, which must be expressed as an IP address
2038 (not a DNS name). The <ref table="Open_vSwitch" column="ssl"/>
2039 column in the <ref table="Open_vSwitch"/> table must point to a
2040 valid SSL configuration when this form is used.
2043 SSL support is an optional feature that is not always built as
2044 part of Open vSwitch.
2048 <dt><code>tcp:<var>ip</var></code>[<code>:<var>port</var></code>]</dt>
2050 The specified TCP <var>port</var> (default: 6632) on the host at
2051 the given <var>ip</var>, which must be expressed as an IP address
2054 <dt><code>pssl:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2057 Listens for SSL connections on the specified TCP <var>port</var>
2058 (default: 6632). If <var>ip</var>, which must be expressed as an
2059 IP address (not a DNS name), is specified, then connections are
2060 restricted to the specified local IP address.
2063 The <ref table="Open_vSwitch" column="ssl"/> column in the <ref
2064 table="Open_vSwitch"/> table must point to a valid SSL
2065 configuration when this form is used.
2068 SSL support is an optional feature that is not always built as
2069 part of Open vSwitch.
2072 <dt><code>ptcp:</code>[<var>port</var>][<code>:<var>ip</var></code>]</dt>
2074 Listens for connections on the specified TCP <var>port</var>
2075 (default: 6632). If <var>ip</var>, which must be expressed as an
2076 IP address (not a DNS name), is specified, then connections are
2077 restricted to the specified local IP address.
2080 <p>When multiple managers are configured, the <ref column="target"/>
2081 values must be unique. Duplicate <ref column="target"/> values yield
2082 unspecified results.</p>
2085 <column name="connection_mode">
2087 If it is specified, this setting must be one of the following strings
2088 that describes how Open vSwitch contacts this OVSDB client over the
2093 <dt><code>in-band</code></dt>
2095 In this mode, this connection's traffic travels over a bridge
2096 managed by Open vSwitch. With this setting, Open vSwitch allows
2097 traffic to and from the client regardless of the contents of the
2098 OpenFlow flow table. (Otherwise, Open vSwitch would never be able
2099 to connect to the client, because it did not have a flow to enable
2100 it.) This is the most common connection mode because it is not
2101 necessary to maintain two independent networks.
2103 <dt><code>out-of-band</code></dt>
2105 In this mode, the client's traffic uses a control network separate
2106 from that managed by Open vSwitch, that is, Open vSwitch does not
2107 use any of its own network devices to communicate with the client.
2108 The control network must be configured separately, before or after
2109 <code>ovs-vswitchd</code> is started.
2114 If not specified, the default is implementation-specific.
2119 <group title="Client Failure Detection and Handling">
2120 <column name="max_backoff">
2121 Maximum number of milliseconds to wait between connection attempts.
2122 Default is implementation-specific.
2125 <column name="inactivity_probe">
2126 Maximum number of milliseconds of idle time on connection to the client
2127 before sending an inactivity probe message. If Open vSwitch does not
2128 communicate with the client for the specified number of seconds, it
2129 will send a probe. If a response is not received for the same
2130 additional amount of time, Open vSwitch assumes the connection has been
2131 broken and attempts to reconnect. Default is implementation-specific.
2132 A value of 0 disables inactivity probes.
2136 <group title="Other Features">
2137 <column name="external_ids">
2138 Key-value pairs for use by external frameworks that integrate with Open
2139 vSwitch, rather than by Open vSwitch itself. System integrators should
2140 either use the Open vSwitch development mailing list to coordinate on
2141 common key-value definitions, or choose key names that are likely to be
2142 unique. No common key-value pairs are currently defined.
2146 <group title="Status">
2147 <column name="is_connected">
2148 <code>true</code> if currently connected to this manager,
2149 <code>false</code> otherwise.
2152 <column name="status">
2153 <p>Key-value pairs that report manager status.</p>
2155 <dt><code>last_error</code></dt>
2156 <dd>A human-readable description of the last error on the connection
2157 to the manager; i.e. <code>strerror(errno)</code>. This key
2158 will exist only if an error has occurred.</dd>
2161 <dt><code>state</code></dt>
2162 <dd>The state of the connection to the manager. Possible values
2163 are: <code>VOID</code> (connection is disabled),
2164 <code>BACKOFF</code> (attempting to reconnect at an increasing
2165 period), <code>CONNECTING</code> (attempting to connect),
2166 <code>ACTIVE</code> (connected, remote host responsive), and
2167 <code>IDLE</code> (remote host idle, sending keep-alive). These
2168 values may change in the future. They are provided only for human
2172 <dt><code>sec_since_connect</code></dt>
2173 <dd>The amount of time since this manager last successfully connected
2174 to the database (in seconds). Value is empty if manager has never
2175 successfully connected.</dd>
2178 <dt><code>sec_since_disconnect</code></dt>
2179 <dd>The amount of time since this manager last disconnected from the
2180 database (in seconds). Value is empty if manager has never
2187 <table name="NetFlow">
2188 A NetFlow target. NetFlow is a protocol that exports a number of
2189 details about terminating IP flows, such as the principals involved
2192 <column name="targets">
2193 NetFlow targets in the form
2194 <code><var>ip</var>:<var>port</var></code>. The <var>ip</var>
2195 must be specified numerically, not as a DNS name.
2198 <column name="engine_id">
2199 Engine ID to use in NetFlow messages. Defaults to datapath index
2203 <column name="engine_type">
2204 Engine type to use in NetFlow messages. Defaults to datapath
2205 index if not specified.
2208 <column name="active_timeout">
2209 The interval at which NetFlow records are sent for flows that are
2210 still active, in seconds. A value of <code>0</code> requests the
2211 default timeout (currently 600 seconds); a value of <code>-1</code>
2212 disables active timeouts.
2215 <column name="add_id_to_interface">
2216 <p>If this column's value is <code>false</code>, the ingress and egress
2217 interface fields of NetFlow flow records are derived from OpenFlow port
2218 numbers. When it is <code>true</code>, the 7 most significant bits of
2219 these fields will be replaced by the least significant 7 bits of the
2220 engine id. This is useful because many NetFlow collectors do not
2221 expect multiple switches to be sending messages from the same host, so
2222 they do not store the engine information which could be used to
2223 disambiguate the traffic.</p>
2224 <p>When this option is enabled, a maximum of 508 ports are supported.</p>
2227 <column name="external_ids">
2228 Key-value pairs for use by external frameworks that integrate with Open
2229 vSwitch, rather than by Open vSwitch itself. System integrators should
2230 either use the Open vSwitch development mailing list to coordinate on
2231 common key-value definitions, or choose key names that are likely to be
2232 unique. No common key-value pairs are currently defined.
2237 SSL configuration for an Open_vSwitch.
2239 <column name="private_key">
2240 Name of a PEM file containing the private key used as the switch's
2241 identity for SSL connections to the controller.
2244 <column name="certificate">
2245 Name of a PEM file containing a certificate, signed by the
2246 certificate authority (CA) used by the controller and manager,
2247 that certifies the switch's private key, identifying a trustworthy
2251 <column name="ca_cert">
2252 Name of a PEM file containing the CA certificate used to verify
2253 that the switch is connected to a trustworthy controller.
2256 <column name="bootstrap_ca_cert">
2257 If set to <code>true</code>, then Open vSwitch will attempt to
2258 obtain the CA certificate from the controller on its first SSL
2259 connection and save it to the named PEM file. If it is successful,
2260 it will immediately drop the connection and reconnect, and from then
2261 on all SSL connections must be authenticated by a certificate signed
2262 by the CA certificate thus obtained. <em>This option exposes the
2263 SSL connection to a man-in-the-middle attack obtaining the initial
2264 CA certificate.</em> It may still be useful for bootstrapping.
2267 <column name="external_ids">
2268 Key-value pairs for use by external frameworks that integrate with Open
2269 vSwitch, rather than by Open vSwitch itself. System integrators should
2270 either use the Open vSwitch development mailing list to coordinate on
2271 common key-value definitions, or choose key names that are likely to be
2272 unique. No common key-value pairs are currently defined.
2276 <table name="sFlow">
2277 <p>An sFlow(R) target. sFlow is a protocol for remote monitoring
2280 <column name="agent">
2281 Name of the network device whose IP address should be reported as the
2282 ``agent address'' to collectors. If not specified, the IP address
2283 defaults to the <ref table="Controller" column="local_ip"/> in the
2284 collector's <ref table="Controller"/>. If an agent IP address cannot be
2285 determined either way, sFlow is disabled.
2288 <column name="header">
2289 Number of bytes of a sampled packet to send to the collector.
2290 If not specified, the default is 128 bytes.
2293 <column name="polling">
2294 Polling rate in seconds to send port statistics to the collector.
2295 If not specified, defaults to 30 seconds.
2298 <column name="sampling">
2299 Rate at which packets should be sampled and sent to the collector.
2300 If not specified, defaults to 400, which means one out of 400
2301 packets, on average, will be sent to the collector.
2304 <column name="targets">
2305 sFlow targets in the form
2306 <code><var>ip</var>:<var>port</var></code>.
2309 <column name="external_ids">
2310 Key-value pairs for use by external frameworks that integrate with Open
2311 vSwitch, rather than by Open vSwitch itself. System integrators should
2312 either use the Open vSwitch development mailing list to coordinate on
2313 common key-value definitions, or choose key names that are likely to be
2314 unique. No common key-value pairs are currently defined.
2318 <table name="Capability">
2319 <p>Records in this table describe functionality supported by the hardware
2320 and software platform on which this Open vSwitch is based. Clients
2321 should not modify this table.</p>
2323 <p>A record in this table is meaningful only if it is referenced by the
2324 <ref table="Open_vSwitch" column="capabilities"/> column in the
2325 <ref table="Open_vSwitch"/> table. The key used to reference it, called
2326 the record's ``category,'' determines the meanings of the
2327 <ref column="details"/> column. The following general forms of
2328 categories are currently defined:</p>
2331 <dt><code>qos-<var>type</var></code></dt>
2332 <dd><var>type</var> is supported as the value for
2333 <ref column="type" table="QoS"/> in the <ref table="QoS"/> table.
2337 <column name="details">
2338 <p>Key-value pairs that describe capabilities. The meaning of the pairs
2339 depends on the category key that the <ref table="Open_vSwitch"
2340 column="capabilities"/> column in the <ref table="Open_vSwitch"/> table
2341 uses to reference this record, as described above.</p>
2343 <p>The presence of a record for category <code>qos-<var>type</var></code>
2344 indicates that the switch supports <var>type</var> as the value of
2345 the <ref table="QoS" column="type"/> column in the <ref table="QoS"/>
2346 table. The following key-value pairs are defined to further describe
2347 QoS capabilities:</p>
2350 <dt><code>n-queues</code></dt>
2351 <dd>Number of supported queues, as a positive integer. Keys in the
2352 <ref table="QoS" column="queues"/> column for <ref table="QoS"/>
2353 records whose <ref table="QoS" column="type"/> value
2354 equals <var>type</var> must range between 0 and this value minus one,
projects / openvswitch / blob