3 - The "ovsdbmonitor" graphical tool has been removed, because it was
4 poorly maintained and not widely used.
5 - New "check-ryu" Makefile target for running Ryu tests for OpenFlow
6 controllers against Open vSwitch. See INSTALL for details.
7 - Added IPFIX support for SCTP flows and templates for ICMPv4/v6 flows.
12 - Address prefix tracking support for flow tables. New columns
13 "prefixes" in OVS-DB table "Flow_Table" controls which packet
14 header fields are used for address prefix tracking. Prefix
15 tracking allows the classifier to skip rules with longer than
16 necessary prefixes, resulting in better wildcarding for datapath
17 flows. Default configuration is to not use any fields for prefix
18 tracking. However, if any flow tables contain both exact matches
19 and masked matches for IP address fields, OVS performance may be
20 increased by using this feature.
21 * As of now, the fields for which prefix lookup can be enabled
22 are: 'tun_id', 'tun_src', 'tun_dst', 'nw_src', 'nw_dst' (or
23 aliases 'ip_src' and 'ip_dst'), 'ipv6_src', and 'ipv6_dst'.
24 (Using this feature for 'tun_id' would only make sense if the
25 tunnel IDs have prefix structure similar to IP addresses.)
26 * There is a maximum number of fields that can be enabled for any
27 one flow table. Currently this limit is 3.
29 $ ovs-vsctl set Bridge br0 flow_tables:0=@N1 -- \
30 --id=@N1 create Flow_Table name=table0
31 $ ovs-vsctl set Bridge br0 flow_tables:1=@N1 -- \
32 --id=@N1 create Flow_Table name=table1
33 $ ovs-vsctl set Flow_Table table0 prefixes=ip_dst,ip_src
34 $ ovs-vsctl set Flow_Table table1 prefixes=[]
35 - TCP flags matching: OVS now supports matching of TCP flags. This
36 has an adverse performance impact when using OVS userspace 1.10
37 or older (no megaflows support) together with the new OVS kernel
38 module. It is recommended that the kernel and userspace modules
39 both are upgraded at the same time.
40 - The default OpenFlow and OVSDB ports will change to
41 IANA-assigned numbers in a future release. Consider updating
42 your installations to specify port numbers instead of using the
45 * The OpenFlow 1.1+ "Write-Actions" instruction is now supported.
46 * OVS limits the OpenFlow port numbers it assigns to port 32767 and
47 below, leaving port numbers above that range free for assignment
49 * ovs-vswitchd now honors changes to the "ofport_request" column
50 in the Interface table by changing the port's OpenFlow port
52 - ovs-vswitchd.conf.db.5 man page will contain graphviz/dot
53 diagram only if graphviz package was installed at the build time.
54 - Support for Linux kernels up to 3.11
56 The "show" command also displays mega flow mask stats.
58 * New command "ofp-parse-pcap" to dump OpenFlow from PCAP files.
59 - ovs-controller has been renamed test-controller. It is no longer
60 packaged or installed by default, because too many users assumed
61 incorrectly that ovs-controller was a necessary or desirable part
62 of an Open vSwitch deployment.
63 - Added vlog option to export to a UDP syslog sink.
65 * The "monitor" command can now monitor all tables in a database,
66 instead of being limited to a single table.
67 - The flow-eviction-threshold has been replaced by the flow-limit which is a
68 hard limit on the number of flows in the datapath. It defaults to 200,000
69 flows. OVS automatically adjusts this number depending on network
71 - Added IPv6 support for active and passive socket communications.
76 - The ovs-vswitchd process is no longer single-threaded. Multiple
77 threads are now used to handle flow set up and asynchronous
80 * Experimental support for OpenFlow 1.1 (in addition to 1.2 and
81 1.3, which had experimental support in 1.10).
82 * Experimental protocol support for OpenFlow 1.1+ groups. This
83 does not yet include an implementation in the Open vSwitch
85 * Experimental protocol support for OpenFlow 1.2+ meters. This
86 does not yet include an implementation in the Open vSwitch
88 * New support for matching outer source and destination IP address
89 of tunneled packets, for tunnel ports configured with the newly
90 added "remote_ip=flow" and "local_ip=flow" options.
91 * Support for matching on metadata 'pkt_mark' for interacting with
92 other system components. On Linux this corresponds to the skb
94 * Support matching, rewriting SCTP ports
95 - The Interface table in the database has a new "ifindex" column to
96 report the interface's OS-assigned ifindex.
97 - New "check-oftest" Makefile target for running OFTest against Open
98 vSwitch. See README-OFTest for details.
99 - The flow eviction threshold has been moved to the Open_vSwitch table.
100 - Database names are now mandatory when specifying ovsdb-server options
101 through database paths (e.g. Private key option with the database name
102 should look like "--private-key=db:Open_vSwitch,SSL,private_key").
103 - Added ovs-dev.py, a utility script helpful for Open vSwitch developers.
104 - Support for Linux kernels up to 3.10
106 * New "ofp-parse" for printing OpenFlow messages read from a file.
107 * New commands for OpenFlow 1.1+ groups.
108 - Added configurable flow caching support to IPFIX exporter.
109 - Dropped support for Linux pre-2.6.32.
110 - Log file timestamps and ovsdb commit timestamps are now reported
111 with millisecond resolution. (Previous versions only reported
115 v1.11.0 - 28 Aug 2013
116 ---------------------
117 - Support for megaflows, which allows wildcarding in the kernel (and
118 any dpif implementation that supports wildcards). Depending on
119 the flow table and switch configuration, flow set up rates are
120 close to the Linux bridge.
121 - The "tutorial" directory contains a new tutorial for some advanced
122 Open vSwitch features.
123 - Stable bond mode has been removed.
124 - The autopath action has been removed.
125 - New support for the data encapsulation format of the LISP tunnel
126 protocol (RFC 6830). An external control plane or manual flow
127 setup is required for EID-to-RLOC mapping.
129 * The "dec_mpls_ttl" and "set_mpls_ttl" actions from OpenFlow
130 1.1 and later are now implemented.
131 * New "stack" extension for use in actions, to push and pop from
133 * The "load" and "set_field" actions can now modify the "in_port". (This
134 allows one to enable output to a flow's input port by setting the
135 in_port to some unused value, such as OFPP_NONE.)
137 * New debugging commands "add-flow", "mod-flow", "del-flow".
138 * "dump-flows" now has a -m option to increase output verbosity.
139 - In dpif-based bridges, cache action translations, which can improve
140 flow set up performance by 80% with a complicated flow table.
141 - New syslog format, prefixed with "ovs|", to be easier to filter.
142 - RHEL: Removes the default firewall rule that allowed GRE traffic to
143 pass through. Any users that relied on this automatic firewall hole
144 will have to manually configure it. The ovs-ctl(8) manpage documents
145 the "enable-protocol" command that can be used as an alternative.
146 - New CFM demand mode which uses data traffic to indicate interface
149 v1.10.0 - 01 May 2013
150 ---------------------
151 - Bridge compatibility support has been removed. Any uses that
152 rely on ovs-brcompatd will have to stick with Open vSwitch 1.9.x
153 or adapt to native Open vSwitch support (e.g. use ovs-vsctl instead
155 - The maximum size of the MAC learning table is now configurable.
156 - With the Linux datapath, packets for new flows are now queued
157 separately on a per-port basis, so it should no longer be
158 possible for a large number of new flows arriving on one port to
159 prevent new flows from being processed on other ports.
161 * Previously ovs-vsctl would retry connecting to the database forever,
162 causing it to hang if ovsdb-server was not running. Now, ovs-vsctl
163 only tries once by default (use --retry to try forever). This change
164 means that you may want to remove uses of --timeout to avoid hangs
166 * Many "ovs-vsctl" database commands now accept an --if-exists option.
167 Please refer to the ovs-vsctl manpage for details.
169 - Experimental support for newer versions of OpenFlow. See
170 the "What versions of OpenFlow does Open vSwitch support?"
171 question in the FAQ for more details.
172 - The OpenFlow "dp_desc" may now be configured by setting the
173 value of other-config:dp-desc in the Bridge table.
174 - It is possible to request the OpenFlow port number with the
175 "ofport_request" column in the Interface table.
176 - The NXM flow_removed message now reports the OpenFlow table ID
177 from which the flow was removed.
179 - New support for the VXLAN tunnel protocol (see the IETF draft here:
180 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03).
181 - Tunneling requires the version of the kernel module paired with
182 Open vSwitch 1.9.0 or later.
183 - Inheritance of the Don't Fragment bit in IP tunnels (df_inherit)
184 is no longer supported.
185 - Path MTU discovery is no longer supported.
186 - CAPWAP tunneling support removed.
187 - Tunnels with multicast destination ports are no longer supported.
189 - The "dump-flows" and "del-flows" no longer require an argument
190 if only one datapath exists.
192 - New "vlog/disable-rate-limit" and "vlog/enable-rate-limit"
193 commands available allow control over logging rate limits.
194 - New "dpif/dump-dps", "dpif/show", and "dpif/dump-flows" command
195 that mimic the equivalent ovs-dpctl commands.
196 - The ofproto library is now responsible for assigning OpenFlow port
197 numbers. An ofproto implementation should assign them when
198 port_construct() is called.
199 - All dpif-based bridges of a particular type share a common
200 datapath called "ovs-<type>", e.g. "ovs-system". The ovs-dpctl
201 commands will now return information on that shared datapath. To
202 get the equivalent bridge-specific information, use the new
203 "ovs-appctl dpif/*" commands.
204 - Backward-incompatible changes:
205 - Earlier Open vSwitch versions treated ANY as a wildcard in flow
206 syntax. OpenFlow 1.1 adds a port named ANY, which introduces a
207 conflict. ANY was rarely used in flow syntax, so we chose to
208 retire that meaning of ANY in favor of the OpenFlow 1.1 meaning.
209 - Patch ports no longer require kernel support, so they now work
210 with FreeBSD and the kernel module built into Linux 3.3 and later.
211 - New "sample" action.
215 ------------------------
217 - Support for ipv6 set action.
218 - SKB mark matching and setting.
219 - support for Linux kernels up to 3.8
220 - FreeBSD is now a supported platform, thanks to code contributions from
221 Gaetano Catalli, Ed Maste, and Giuseppe Lettieri.
222 - ovs-bugtool: New --ovs option to report only OVS related information.
223 - New %t and %T log escapes to identify the subprogram within a
224 cooperating group of processes or threads that emitted a log message.
225 The default log patterns now include this information.
227 - Allow bitwise masking for SHA and THA fields in ARP, SLL and TLL
228 fields in IPv6 neighbor discovery messages, and IPv6 flow label.
229 - Adds support for writing to the metadata field for a flow.
231 - The tunneling code no longer assumes input and output keys are
232 symmetric. If they are not, PMTUD needs to be disabled for
233 tunneling to work. Note this only applies to flow-based keys.
234 - New support for a nonstandard form of GRE that supports a 64-bit key.
235 - Tunnel Path MTU Discovery default value was set to 'disabled'.
236 This feature is deprecated and will be removed soon.
237 - Tunnel header caching removed.
239 - Commands and actions that accept port numbers now also accept keywords
240 that represent those ports (such as LOCAL, NONE, and ALL). This is
241 also the recommended way to specify these ports, for compatibility
242 with OpenFlow 1.1 and later (which use the OpenFlow 1.0 numbers
243 for these ports for different purposes).
245 - Support requesting the port number with the "port_no" option in
246 the "add-if" command.
247 - ovs-pki: The "online PKI" features have been removed, along with
248 the ovs-pki-cgi program that facilitated it, because of some
249 alarmist insecurity claims. We do not believe that these claims
250 are true, but because we do not know of any users for this
251 feature it seems better on balance to remove it. (The ovs-pki-cgi
252 program was not included in distribution packaging.)
253 - ovsdb-server now enforces the immutability of immutable columns. This
254 was not enforced in earlier versions due to an oversight.
255 - The following features are now deprecated. They will be removed no
256 earlier than February 2013. Please email dev@openvswitch.org with
258 - Bridge compatibility.
260 - The autopath action.
261 - Interface type "null".
262 - Numeric values for reserved ports (see "ovs-ofctl" note above).
263 - Tunnel Path MTU Discovery.
264 - CAPWAP tunnel support.
265 - The data in the RARP packets can now be matched in the same way as the
270 ------------------------
271 *** Internal only release ***
272 - New FAQ. Please send updates and additions!
273 - Authors of controllers, please read the new section titled "Action
274 Reproduction" in DESIGN, which describes an Open vSwitch change in
275 behavior in corner cases that may affect some controllers.
277 - A new test utility that can create L3 tunnel between two Open
278 vSwitches and detect connectivity issues.
280 - New --sort and --rsort options for "dump-flows" command.
281 - "mod-port" command can now control all OpenFlow config flags.
283 - Allow general bitwise masking for IPv4 and IPv6 addresses in
284 IPv4, IPv6, and ARP packets. (Previously, only CIDR masks
286 - Allow support for arbitrary Ethernet masks. (Previously, only
287 the multicast bit in the destination address could be individually
289 - New field OXM_OF_METADATA, to align with OpenFlow 1.1.
290 - The OFPST_QUEUE request now reports an error if a specified port or
291 queue does not exist, or for requests for a specific queue on all
292 ports, if the specified queue does not exist on any port. (Previous
293 versions generally reported an empty set of results.)
294 - New "flow monitor" feature to allow controllers to be notified of
295 flow table changes as they happen.
296 - Additional protocols are not mirrored and dropped when forward-bpdu is
297 false. For a full list, see the ovs-vswitchd.conf.db man page.
298 - Open vSwitch now sends RARP packets in situations where it previously
299 sent a custom protocol, making it consistent with behavior of QEMU and
301 - All Open vSwitch programs and log files now show timestamps in UTC,
302 instead the local timezone, by default.
306 ------------------------
307 - kernel modules are renamed. openvswitch_mod.ko is now
308 openvswitch.ko and brcompat_mod.ko is now brcompat.ko.
309 - Increased the number of NXM registers to 8.
310 - Added ability to configure DSCP setting for manager and controller
311 connections. By default, these connections have a DSCP value of
312 Internetwork Control (0xc0).
313 - Added the granular link health statistics, 'cfm_health', to an
316 - Added support to mask nd_target for ICMPv6 neighbor discovery flows.
317 - Added support for OpenFlow 1.3 port description (OFPMP_PORT_DESC)
320 - Added the "dump-ports-desc" command to retrieve port
321 information using the new port description multipart messages.
323 - Added support for spawning ovs-test server from the client.
324 - Now ovs-test is able to automatically create test bridges and ports.
325 - "ovs-dpctl dump-flows" now prints observed TCP flags in TCP flows.
326 - Tripled flow setup performance.
327 - The "coverage/log" command previously available through ovs-appctl
328 has been replaced by "coverage/show". The new command replies with
329 coverage counter values, instead of logging them.
333 ------------------------
334 - Allow OFPP_CONTROLLER as the in_port for packet-out messages.
338 ------------------------
339 *** Internal only release ***
341 - LACP bonds no longer fall back to balance-slb when negotiations fail.
342 Instead they drop traffic.
343 - The default bond_mode changed from SLB to active-backup, to protect
344 unsuspecting users from the significant risks of SLB bonds (which are
345 documented in vswitchd/INTERNALS).
346 - Load balancing can be disabled by setting the bond-rebalance-interval
349 - Added support for bitwise matching on TCP and UDP ports.
350 See ovs-ofctl(8) for more information.
351 - NXM flow dumps now include times elapsed toward idle and hard
353 - Added an OpenFlow extension NXT_SET_ASYNC_CONFIG that allows
354 controllers more precise control over which OpenFlow messages they
355 receive asynchronously.
356 - New "fin_timeout" action.
357 - Added "fin_timeout" support to "learn" action.
358 - New Nicira action NXAST_CONTROLLER that offers additional features
359 over output to OFPP_CONTROLLER.
360 - When QoS settings for an interface do not configure a default queue
361 (queue 0), Open vSwitch now uses a default configuration for that
362 queue, instead of dropping all packets as in previous versions.
364 - Logging to console and file will have UTC timestamp as a default for
365 all the daemons. An example of the default format is
366 2012-01-27T16:35:17Z. ovs-appctl can be used to change the default
368 - The syntax of commands and options to set log levels was simplified,
369 to make it easier to remember.
370 - New support for limiting the number of flows in an OpenFlow flow
371 table, with configurable policy for evicting flows upon
372 overflow. See the Flow_Table table in ovs-vswitch.conf.db(5)
373 for more information.
374 - New "enable-async-messages" column in the Controller table. If set to
375 false, OpenFlow connections to the controller will initially have all
376 asynchronous messages disabled, overriding normal OpenFlow behavior.
377 - ofproto-provider interface:
378 - "struct rule" has a new member "used" that ofproto implementations
379 should maintain by updating with ofproto_rule_update_used().
381 - The new option --timestamp causes the "monitor" command to print
382 a timestamp with every update.
383 - CFM module CCM broadcasts can now be tagged with an 802.1p priority.
387 ------------------------
389 - Added support for querying, modifying, and deleting flows
390 based on flow cookie when using NXM.
391 - Added new NXM_PACKET_IN format.
392 - Added new NXAST_DEC_TTL action.
394 - Added daemonization support to the monitor and snoop commands.
396 - The "find" command supports new set relational operators
397 {=}, {!=}, {<}, {>}, {<=}, and {>=}.
398 - ovsdb-tool now uses the typical database and schema installation
399 directories as defaults.
400 - The default MAC learning timeout has been increased from 60 seconds
401 to 300 seconds. The MAC learning timeout is now configurable.
405 ------------------------
406 - Compatible with Open vSwitch kernel module included in Linux 3.3.
407 - New "VLAN splinters" feature to work around buggy device drivers
408 in old Linux versions. (This feature is deprecated. When
409 broken device drivers are no longer in widespread use, we will
410 delete this feature.) See ovs-vswitchd.conf.db(5) for more
413 - Added ability to match on IPv6 flow label through NXM.
414 - Added ability to match on ECN bits in IPv4 and IPv6 through NXM.
415 - Added ability to match on TTL in IPv4 and IPv6 through NXM.
416 - Added ability to modify ECN bits in IPv4.
417 - Added ability to modify TTL in IPv4.
419 - Don't require the "normal" action to use mirrors. Traffic will
420 now be properly mirrored for any flows, regardless of their
422 - Track packet and byte statistics sent on mirrors.
423 - The sFlow implementation can now usually infer the correct agent
424 device instead of having to be told explicitly.
426 - New "fdb/flush" command to flush bridge's MAC learning table.
428 - A new distributed testing tool that allows one to diagnose performance
429 and connectivity issues. This tool currently is not included in RH or
431 - RHEL packaging now supports integration with Red Hat network scripts.
433 - Post 1.4.*, OVS will be changing the default bond mode from balance-slb
434 to active-backup. SLB bonds carry significant risks with them
435 (documented vswitchd/INTERNALS) which we want to prevent unsuspecting
436 users from running into. Users are advised to update any scripts or
437 configuration which may be negatively impacted by explicitly setting
438 the bond mode which they want to use.
442 ------------------------
444 - Added an OpenFlow extension which allows the "output" action to accept
446 - Added an OpenFlow extension for flexible learning.
447 - Bumped number of NXM registers from four to five.
449 - New "version" command to determine version of running daemon.
450 - If no argument is provided for "cfm/show", displays detailed
451 information about all interfaces with CFM enabled.
452 - If no argument is provided for "lacp/show", displays detailed
453 information about all ports with LACP enabled.
455 - New "set-if" command to modify a datapath port's configuration.
457 - The software switch now supports 255 OpenFlow tables, instead
458 of just one. By default, only table 0 is consulted, but the
459 new NXAST_RESUBMIT_TABLE action can look up in additional
460 tables. Tables 128 and above are reserved for use by the
461 switch itself; please use only tables 0 through 127.
462 - Add support for 802.1D spanning tree (STP).
463 - Fragment handling extensions:
464 - New OFPC_FRAG_NX_MATCH fragment handling mode, in which L4
465 fields are made available for matching in fragments with
467 - New NXM_NX_IP_FRAG match field for matching IP fragments (usable
468 via "ip_frag" in ovs-ofctl).
469 - New ovs-ofctl "get-frags" and "set-frags" commands to get and set
470 fragment handling policy.
471 - CAPWAP tunneling now supports an extension to transport a 64-bit key.
472 By default it remains compatible with the old version and other
473 standards-based implementations.
474 - Flow setups are now processed in a round-robin manner across ports
475 to prevent any single client from monopolizing the CPU and conducting
476 a denial of service attack.
477 - Added support for native VLAN tagging. A new "vlan_mode"
478 parameter can be set for "port". Possible values: "access",
479 "trunk", "native-tagged" and "native-untagged".
480 - test-openflowd has been removed. Please use ovs-vswitchd instead.
483 ------------------------
484 - New "ofproto" abstraction layer to ease porting to hardware
486 - Packaging for Red Hat Enterprise Linux 5.6 and 6.0.
487 - Datapath support for Linux kernels up to 3.0.
489 - New "bundle" and "bundle_load" action extensions.
491 - Implement table unique constraints.
492 - Support cooperative locking between callers.
494 - New "-s" option for "show" command prints packet and byte
495 counters for each port.
497 - New "--readd" option for "replace-flows".
499 - New "show" command to print an overview of configuration.
500 - New "comment" command to add remark that explains intentions.
501 - ovs-brcompatd has been rewritten to fix long-standing bugs.
502 - ovs-openflowd has been renamed test-openflowd and moved into the
503 tests directory. Its presence confused too many users. Please
504 use ovs-vswitchd instead.
505 - New ovs-benchmark utility to test flow setup performance.
506 - A new log level "off" has been added. Configuring a log facility
507 "off" prevents any messages from being logged to it. Previously,
508 "emer" was effectively "off" because no messages were ever logged at
509 level "emer". Now, errors that cause a process to exit are logged
511 - "configure" option --with-l26 has been renamed --with-linux, and
512 --with-l26-source has been renamed --with-linux-source. The old
513 names will be removed after the next release, so please update
515 - The "-2.6" suffix has been dropped from the datapath/linux-2.6 and
516 datapath/linux-2.6/compat-2.6 directories.
518 - Dropped support for "tun_id_from_cookie" OpenFlow extension.
519 Please use the extensible match extensions instead.
520 - Removed the Maintenance_Point and Monitor tables in an effort
521 to simplify 802.1ag configuration.
522 - Performance and scalability improvements
526 ------------------------
527 - Ability to define policies over IPv6
530 - Support for extensible match extensions to OpenFlow
532 - Support for HFSC qdisc.
533 - Queue used by in-band control can now be configured.
535 - Kernel<->userspace interface has been reworked and should be
536 close to a stable ABI now.
537 - "Port group" concept has been dropped.
538 - GRE over IPSEC tunnels
540 - New active backup bonding mode.
541 - New L4 hashing support when LACP is enabled.
542 - Source MAC hash now includes VLAN field also.
544 - Greatly improved handling of large flow tables
546 - "show" command now prints full vport configuration.
547 - "dump-groups" command removed since kernel support for
548 port groups was dropped.
550 - New commands for working with the new Managers table.
551 - "list" command enhanced with new formatting options and --columns
553 - "get" command now accepts new --id option.
554 - New "find" command.
556 - New "queue-stats" command for printing queue stats.
557 - New commands "replace-flows" and "diff-flows".
558 - Commands to add and remove flows can now read from files.
559 - New --flow-format option to enable or disable NXM.
560 - New --more option to increase OpenFlow message verbosity.
561 - Removed "tun-cookie" command, which is no longer useful.
562 - ovs-controller enhancements for testing various features.
563 - New ovs-vlan-test command for testing for Linux kernel driver VLAN
564 bugs. New ovs-vlan-bug-workaround command for enabling and
565 disabling a workaround for these driver bugs.
567 - "Resubmit" actions now update flow statistics.
568 - New "register" extension for use in matching and actions, via NXM.
569 - New "multipath" experimental action extension.
570 - New support for matching multicast Ethernet frames, via NXM.
571 - New extension for OpenFlow vendor error codes.
572 - New extension to set the QoS output queue without actually
573 sending to an output port.
574 - Open vSwitch now reports a single flow table, instead of
575 separate hash and wildcard tables. This better models the
576 current implementation.
577 - New experimental "note" action.
578 - New "ofproto/trace" ovs-appctl command and associated utilities
579 to ease debugging complex flow tables.
581 - Schema documentation now includes an entity-relationship diagram.
582 - The database is now garbage collected. In most tables,
583 unreferenced rows will be deleted automatically.
584 - Many tables now include statistics updated periodically by
585 ovs-vswitchd or ovsdb-server.
586 - Every table now has an "external-ids" column for use by OVS
588 - There is no default controller anymore. Each bridge must have its
589 controller individually specified.
590 - The "fail-mode" is now a property of a Bridge instead of a Controller.
591 - New versioning and checksum features.
592 - New Managers table and manager_options column in Open_vSwitch table
593 for specifying managers. The old "managers" column in the
594 Open_vSwitch table has been removed.
595 - Many "name" columns are now immutable.
597 - Dropped support for XenServer pre-5.6.100.
598 - Dropped support for Linux pre-2.6.18.
599 - Dropped controller discovery support.
600 - Dropped "ovs-ofctl status" and the OpenFlow extension that it used.
601 Statistics reporting in the database is a rough equivalent.
602 - Dropped the "corekeeper" package (now separate, at
603 http://openvswitch.org/cgi-bin/gitweb.cgi?p=corekeeper).
604 - Performance and scalability improvements
607 v1.1.0pre2 - 13 Sep 2010
608 ------------------------
611 v1.1.0pre1 - 31 Aug 2010
612 ------------------------
613 - OpenFlow 1.0 slicing (QoS) functionality
614 - Python bindings for configuration database (no write support)
615 - Performance and scalability improvements
620 - New "patch" interface type
625 - Configuration database with remote management
628 - Support for XenServer 5.5 and 5.6
629 - Performance and scalability improvements
632 v0.99.2 - 18 Feb 2010
633 ---------------------
636 v0.99.1 - 25 Jan 2010
637 ---------------------
638 - Add support for sFlow(R)
639 - Make headers compatible with C++
642 v0.99.0 - 14 Jan 2010
643 ---------------------
644 - User-space forwarding engine
647 v0.90.7 - 29 Nov 2009
648 ---------------------
649 - Add support for NetFlow active timeouts
656 v0.90.5 - 21 Sep 2009
657 ---------------------
658 - Generalize in-band control to more diverse network setups