4 * The "requested-chassis" option for a logical switch port now accepts a
5 chassis "hostname" in addition to a chassis "name".
10 * ovs-ofctl can now accept and display port names in place of numbers. By
11 default it always accepts names and in interactive use it displays them;
12 use --names or --no-names to override. See ovs-ofctl(8) for details.
13 * "ovs-ofctl dump-flows" now accepts --no-stats to omit flow statistics.
14 - New ovs-dpctl command "ct-stats-show" to show connection tracking stats.
16 * Added support to set packet mark for tunnel endpoint using
17 `egress_pkt_mark` OVSDB option.
18 * When using Linux kernel datapath tunnels may be created using rtnetlink.
19 This will allow us to take advantage of new tunnel features without
20 having to make changes to the vport modules.
21 - EMC insertion probability is reduced to 1% and is configurable via
22 the new 'other_config:emc-insert-inv-prob' option.
24 * DPDK log messages redirected to OVS logging subsystem.
25 Log level can be changed in a usual OVS way using
26 'ovs-appctl vlog' commands for 'dpdk' module. Lower bound
27 still can be configured via extra arguments for DPDK EAL.
28 * dpdkvhostuser ports are marked as deprecated. They will be removed
29 in an upcoming release.
30 * Support for DPDK v17.05.1.
31 - IPFIX now provides additional counters:
32 * Total counters since metering process startup.
33 * Per-flow TCP flag counters.
34 * Multicast, broadcast, and unicast counters.
35 - New support for multiple VLANs (802.1ad or "QinQ"), including a new
36 "dot1q-tunnel" port VLAN mode.
37 - In ovn-vsctl and vtep-ctl, record UUIDs in commands may now be
38 abbreviated to 4 hex digits.
40 * Added NAT support for userspace datapath.
41 * Added FTP and TFTP support with NAT for userspace datapath.
42 * Experimental NSH (Network Service Header) support in userspace datapath.
44 * New built-in DNS support.
45 * IPAM for IPv4 can now exclude user-defined addresses from assignment.
46 * IPAM can now assign IPv6 addresses.
47 * Make the DHCPv4 router setting optional.
48 * Gratuitous ARP for NAT addresses on a distributed logical router.
49 * Allow ovn-controller SSL configuration to be obtained from vswitchd
51 * ovn-trace now has basic support for tracing distributed firewalls.
52 * In ovn-nbctl and ovn-sbctl, record UUIDs in commands may now be
53 abbreviated to 4 hex digits.
54 * "ovn-sbctl lflow-list" can now print OpenFlow flows that correspond
56 * Now uses OVSDB RBAC support to reduce impact of compromised hypervisors.
57 * Multiple chassis may now be specified for L3 gateways. When more than
58 one chassis is specified, OVN will manage high availability for that
60 * Add support for ACL logging.
61 * ovn-northd now has native support for active-standby high availability.
62 - Tracing with ofproto/trace now traces through recirculation.
64 * New support for role-based access control (see ovsdb-server(1)).
65 - New commands 'stp/show' and 'rstp/show' (see ovs-vswitchd(8)).
67 * All features required by OpenFlow 1.4 are now implemented, so
68 ovs-vswitchd now enables OpenFlow 1.4 by default (in addition to
70 * Increased support for OpenFlow 1.6 (draft).
71 * Bundles now support hashing by just nw_src or nw_dst.
72 * The "learn" action now supports a "limit" option (see ovs-ofctl(8)).
73 * The port status bit OFPPS_LIVE now reflects link aliveness.
74 * OpenFlow 1.5 packet-out is now supported.
75 * Support for OpenFlow 1.5 field packet_type and packet-type-aware
77 * Added generic encap and decap actions (EXT-382).
78 First supported use case is encap/decap for Ethernet.
79 * Added NSH (Network Service Header) support in userspace
80 Used generic encap and decap actions to implement encapsulation and
81 decapsulation of NSH header.
82 IETF NSH draft - https://datatracker.ietf.org/doc/draft-ietf-sfc-nsh/
83 * Conntrack state is only available to the processing path that
84 follows the "recirc_table" argument of the ct() action. Starting
85 in OVS 2.8, this state is now cleared for the current processing
86 path whenever ct() is called.
88 * OVN services are no longer restarted automatically after upgrade.
89 * ovs-vswitchd and ovsdb-server run as non-root users by default.
90 - Add --cleanup option to command 'ovs-appctl exit' (see ovs-vswitchd(8)).
92 * Use new tunnel port option "packet_type" to configure L2 vs. L3.
93 * In conjunction with PTAP tunnel ports can handle a mix of L2 and L3
95 * New vxlan tunnel extension "gpe" to support VXLAN-GPE tunnels.
96 * New support for non-Ethernet (L3) payloads in GRE and VXLAN-GPE.
97 - The BFD detection multiplier is now user-configurable.
98 - Add experimental support for hardware offloading
99 * HW offloading is disabled by default.
100 * HW offloading is done through the TC interface.
101 - IPv6 link local addresses are now supported on Linux. Use % to designate
105 ---------------------
106 - Utilities and daemons that support SSL now allow protocols and
107 ciphers to be configured with --ssl-protocols and --ssl-ciphers.
109 * QoS is now implemented via egress shaping rather than ingress policing.
110 * DSCP marking is now supported, via the new northbound QoS table.
111 * IPAM now supports fixed MAC addresses.
112 * Support for source IP address based routing.
114 - New --ovs option to also print OpenFlow flows.
115 - put_dhcp_opts and put_dhcp_optsv6 actions may now be traced.
116 * Support for managing SSL and remote connection configuration in
117 northbound and southbound databases.
118 * TCP connections to northbound and southbound databases are no
119 longer enabled by default and must be explicitly configured.
120 See documentation for ovn-sbctl/ovn-nbctl "set-connection"
121 command or the ovn-ctl "--db-sb-create-insecure-remote" and
122 "--db-nb-create-insecure-remote" command-line options for
123 information regarding remote connection configuration.
124 * New appctl "inject-pkt" command in ovn-controller that allows
125 packets to be injected into the connected OVS instance.
126 * Distributed logical routers may now be connected directly to
127 logical switches with localnet ports, by specifying a
128 "redirect-chassis" on the distributed gateway port of the
129 logical router. NAT rules may be specified directly on the
130 distributed logical router, and are handled either centrally on
131 the "redirect-chassis", or in many cases are handled locally on
132 the hypervisor where the corresponding logical port resides.
133 Gratuitous ARP for NAT addresses on a distributed logical
134 router is not yet supported, but will be added in a future
136 - Fixed regression in table stats maintenance introduced in OVS
137 2.3.0, wherein the number of OpenFlow table hits and misses was
140 * OFPT_PACKET_OUT messages are now supported in bundles.
141 * A new "selection_method=dp_hash" type for OpenFlow select group
142 bucket selection that uses the datapath computed 5-tuple hash
143 without making datapath flows match the 5-tuple fields, which
144 is useful for more efficient load balancing, for example. This
145 uses the Netronome extension to OpenFlow 1.5+ that allows
146 control over the OpenFlow select groups selection method. See
147 "selection_method" and related options in ovs-ofctl(8) for
149 * The "sample" action now supports "ingress" and "egress" options.
150 * The "ct" action now supports the TFTP ALG where support is available.
151 * New actions "clone" and "ct_clear".
152 * The "meter" action is now supported in the userspace datapath.
154 * 'bundle' command now supports packet-out messages.
155 * New syntax for 'ovs-ofctl packet-out' command, which uses the
156 same string parser as the 'bundle' command. The old 'packet-out'
157 syntax is deprecated and will be removed in a later OVS
159 * New unixctl "ofctl/packet-out" command, which can be used to
160 instruct a flow monitor to issue OpenFlow packet-out messages.
162 * Remote connections can now be made read-only (see ovsdb-server(1)).
164 * TLV mappings for protocols such as Geneve are now segregated on
165 a per-OpenFlow bridge basis rather than globally. (The interface
167 * Removed support for IPsec tunnels.
169 * New option 'n_rxq_desc' and 'n_txq_desc' fields for DPDK interfaces
170 which set the number of rx and tx descriptors to use for the given port.
171 * Support for DPDK v16.11.
172 * Support for rx checksum offload. Refer DPDK HOWTO for details.
173 * Port Hotplug is now supported.
174 * DPDK physical ports can now have arbitrary names. The PCI address of
175 the device must be set using the 'dpdk-devargs' option. Compatibility
176 with the old dpdk<portid> naming scheme is broken, and as such a
177 device will not be available for use until a valid dpdk-devargs is
179 * Virtual DPDK Poll Mode Driver (vdev PMD) support.
180 * Removed experimental tag.
182 * A package upgrade does not automatically restart OVS service.
183 - ovs-vswitchd/ovs-vsctl:
184 * Ports now have a "protected" flag. Protected ports can not forward
185 frames to other protected ports. Unprotected ports can receive and
186 forward frames to protected and other unprotected ports.
187 - ovs-vsctl, ovn-nbctl, ovn-sbctl, vtep-ctl:
188 * Database commands now accept integer ranges, e.g. "set port
189 eth0 trunks=1-10" to enable trunking VLANs 1 to 10.
192 ---------------------
193 - First supported release of OVN. See ovn-architecture(7) for more
196 * New "monitor_cond" "monitor_cond_update" and "update2" extensions to
199 * OpenFlow 1.3+ bundles now expire after 10 seconds since the
200 last time the bundle was either opened, modified, or closed.
201 * OpenFlow 1.3 Extension 230, adding OpenFlow Bundles support, is
203 * OpenFlow 1.3+ bundles are now supported for group mods as well as
204 flow mods and port mods. Both 'atomic' and 'ordered' bundle
205 flags are supported for group mods as well as flow mods.
206 * Internal OpenFlow rule representation for load and set-field
207 actions is now much more memory efficient. For a complex flow
208 table this can reduce rule memory consumption by 40%.
209 * Bundles are now much more memory efficient than in OVS 2.5.
210 Together with memory efficiency improvements in OpenFlow rule
211 representation, the peak OVS resident memory use during a
212 bundle commit for large complex set of flow mods can be only
213 25% of that in OVS 2.5 (4x lower).
214 * OpenFlow 1.1+ OFPT_QUEUE_GET_CONFIG_REQUEST now supports OFPP_ANY.
215 * OpenFlow 1.4+ OFPMP_QUEUE_DESC is now supported.
216 * OpenFlow 1.4+ OFPT_TABLE_STATUS is now supported.
217 * New property-based packet-in message format NXT_PACKET_IN2 with support
218 for arbitrary user-provided data and for serializing flow table
219 traversal into a continuation for later resumption.
220 * New extension message NXT_SET_ASYNC_CONFIG2 to allow OpenFlow 1.4-like
221 control over asynchronous messages in earlier versions of OpenFlow.
222 * New OpenFlow extension NXM_NX_MPLS_TTL to provide access to MPLS TTL.
223 * New output option, output(port=N,max_len=M), to allow truncating a
224 packet to size M bytes when outputting to port N.
225 * New command OFPGC_ADD_OR_MOD for OFPT_GROUP_MOD message that adds a
226 new group or modifies an existing groups
227 * The optional OpenFlow packet buffering feature is deprecated in
228 this release, and will be removed in the next OVS release
229 (2.7). After the change OVS always sends the 'buffer_id' as
230 0xffffffff in packet-in messages and will send an error
231 response if any other value of this field is included in
232 packet-out and flow mod sent by a controller. Controllers are
233 already expected to work properly in cases where the switch can
234 not buffer packets, so this change should not affect existing
236 * New OpenFlow extension NXT_CT_FLUSH_ZONE to flush conntrack zones.
237 - Improved OpenFlow version compatibility for actions:
238 * New OpenFlow extension to support the "group" action in OpenFlow 1.0.
239 * OpenFlow 1.0 "enqueue" action now properly translated to OpenFlow 1.1+.
240 * OpenFlow 1.1 "mod_nw_ecn" and OpenFlow 1.1+ "mod_nw_ttl" actions now
241 properly translated to OpenFlow 1.0.
243 * queue-get-config command now allows a queue ID to be specified.
244 * '--bundle' option can now be used with OpenFlow 1.3 and with group mods.
245 * New "bundle" command allows executing a mixture of flow and group mods
246 as a single atomic transaction.
247 * New option "--color" to produce colorized output for some commands.
248 * New option '--may-create' to use OFPGC_ADD_OR_MOD in mod-group command.
250 * New "sampling_port" option for "sample" action to allow sampling
251 ingress and egress tunnel metadata with IPFIX.
252 * New ovs-ofctl commands "dump-ipfix-bridge" and "dump-ipfix-flow" to
253 dump bridge IPFIX statistics and flow based IPFIX statistics.
254 * New setting other-config:virtual_obs_id to add an arbitrary string
257 * OVS Linux datapath now implements Conntrack NAT action with all
258 supported Linux kernels.
259 * Support for truncate action.
260 * New QoS type "linux-noop" that prevents Open vSwitch from trying to
261 manage QoS for a given port (useful when other software manages QoS).
263 * New option "n_rxq" for PMD interfaces.
264 Old 'other_config:n-dpdk-rxqs' is no longer supported.
265 Not supported by vHost interfaces. For them number of rx and tx queues
266 is applied from connected virtio device.
267 * New 'other_config:pmd-rxq-affinity' field for PMD interfaces, that
268 allows to pin port's rx queues to desired cores.
269 * New appctl command 'dpif-netdev/pmd-rxq-show' to check the port/rxq
271 * Type of log messages from PMD threads changed from INFO to DBG.
272 * QoS functionality with sample egress-policer implementation.
273 * The mechanism for configuring DPDK has changed to use database
274 * Sensible defaults have been introduced for many of the required
275 configuration options
276 * DB entries have been added for many of the DPDK EAL command line
277 arguments. Additional arguments can be passed via the dpdk-extra
279 * Add ingress policing functionality.
280 * PMD threads servicing vHost User ports can now come from the NUMA
281 node that device memory is located on if CONFIG_RTE_LIBRTE_VHOST_NUMA
283 * Basic connection tracking for the userspace datapath (no ALG,
284 fragmentation or NAT support yet)
285 * Support for DPDK 16.07
286 * Optional support for DPDK pdump enabled.
287 * Jumbo frame support
288 * Remove dpdkvhostcuse port type.
289 * OVS client mode for vHost and vHost reconnect (Requires QEMU 2.7)
290 * 'dpdkvhostuserclient' port type.
291 - Increase number of registers to 16.
292 - ovs-benchmark: This utility has been removed due to lack of use and
295 * New "vlog/close" command.
297 * Added the ability to selectively start the forwarding and database
298 functions (ovs-vswitchd and ovsdb-server, respectively).
300 * Remove max number of sessions limit, to enable connection scaling
303 * Added support for Python 3.4+ in addition to existing support
306 * Introduced SELinux policy package.
307 - Datapath Linux kernel compatibility.
308 * Dropped support for kernel older than 3.10.
309 * Removed VLAN splinters feature.
310 * Datapath supports kernel upto 4.7.
312 * Flow based tunnel match and action can be used for IPv6 address using
313 tun_ipv6_src, tun_ipv6_dst fields.
314 * Added support for IPv6 tunnels, for details checkout FAQ.
315 * Deprecated support for IPsec tunnels ports.
316 - A wrapper script, 'ovs-tcpdump', to easily port-mirror an OVS port and
318 - Introduce --no-self-confinement flag that allows daemons to work with
319 sockets outside their run directory.
320 - ovs-pki: Changed message digest algorithm from SHA-1 to SHA-512 because
321 SHA-1 is no longer secure and some operating systems have started to
322 disable it in OpenSSL.
323 - Add 'mtu_request' column to the Interface table. It can be used to
324 configure the MTU of the ports.
327 - Using openvswitch module in conjunction with upstream Linux tunnels:
328 * When using the openvswitch module distributed with OVS against kernel
329 versions 4.4 to 4.6, the openvswitch module cannot be loaded or used at
330 the same time as "ip_gre".
331 - Conntrack FTP ALGs: When using the openvswitch module distributed with
332 OVS, particular Linux distribution kernels versions may provide diminished
333 functionality. This typically affects active FTP data connections when
334 using "actions=ct(alg=ftp),..." in flow tables. Specifically:
335 * Centos 7.1 kernels (3.10.0-2xx) kernels are unable to correctly set
336 up expectations for FTP data connections in multiple zones,
337 eg "actions=ct(zone=1,alg=ftp),ct(zone=2,alg=ftp),...". Executing the
338 "ct" action for subsequent data connections may fail to determine that
339 the data connection is "related" to an existing connection.
340 * Centos 7.2 kernels (3.10.0-3xx) kernels may not establish FTP ALG state
341 correctly for NATed connections. As a result, flows that perform NAT,
342 eg "actions=ct(nat,ftp=alg,table=1),..." may fail to NAT the packet,
343 and will populate the "ct_state=inv" bit in the flow.
347 ---------------------
348 - Dropped support for Python older than version 2.7. As a consequence,
349 using Open vSwitch 2.5 or later on XenServer 6.5 or earlier (which
350 have Python 2.4) requires first installing Python 2.7.
352 * Group chaining (where one OpenFlow group triggers another) is
354 * OpenFlow 1.4+ "importance" is now considered for flow eviction.
355 * OpenFlow 1.4+ OFPTC_EVICTION is now implemented.
356 * OpenFlow 1.4+ OFPTC_VACANCY_EVENTS is now implemented.
357 * OpenFlow 1.4+ OFPMP_TABLE_DESC is now implemented.
358 * Allow modifying the ICMPv4/ICMPv6 type and code fields.
359 * OpenFlow 1.4+ OFPT_SET_ASYNC_CONFIG and OFPT_GET_ASYNC_CONFIG are
362 * New "out_group" keyword for OpenFlow 1.1+ matching on output group.
364 * Geneve tunnels can now match and set options and the OAM bit.
365 * The nonstandard GRE64 tunnel extension has been dropped.
366 - Support Multicast Listener Discovery (MLDv1 and MLDv2).
367 - Add 'symmetric_l3l4' and 'symmetric_l3l4+udp' hash functions.
368 - sFlow agent now reports tunnel and MPLS structures.
369 - New 'check-system-userspace', 'check-kmod' and 'check-kernel' Makefile
370 targets to run a new system testsuite. These tests can be run inside
371 a Vagrant box. See INSTALL.md for details
372 - Mark --syslog-target argument as deprecated. It will be removed in
373 the next OVS release.
374 - Added --user option to all daemons
375 - Add support for connection tracking through the new "ct" action
376 and "ct_state"/"ct_zone"/"ct_mark"/"ct_label" match fields. Only
377 available on Linux kernels with the connection tracking module loaded.
378 - Add experimental version of OVN. OVN, the Open Virtual Network, is a
379 system to support virtual network abstraction. OVN complements the
380 existing capabilities of OVS to add native support for virtual network
381 abstractions, such as virtual L2 and L3 overlays and security groups.
383 * DPDK ports may now be created via network scripts (see README.RHEL).
386 * Added multiqueue support to vhost-user
387 * Note: QEMU 2.5+ required for multiqueue support
390 ---------------------
391 - Flow table modifications are now atomic, meaning that each packet
392 now sees a coherent version of the OpenFlow pipeline. For
393 example, if a controller removes all flows with a single OpenFlow
394 "flow_mod", no packet sees an intermediate version of the OpenFlow
395 pipeline where only some of the flows have been deleted.
396 - Added support for SFQ, FQ_CoDel and CoDel qdiscs.
397 - Add bash command-line completion support for ovs-vsctl Please check
398 utilities/ovs-command-compgen.INSTALL.md for how to use.
399 - The MAC learning feature now includes per-port fairness to mitigate
400 MAC flooding attacks.
401 - New support for a "conjunctive match" OpenFlow extension, which
402 allows constructing OpenFlow matches of the form "field1 in
403 {a,b,c...} AND field2 in {d,e,f...}" and generalizations. For details,
404 see documentation for the "conjunction" action in ovs-ofctl(8).
405 - Add bash command-line completion support for ovs-appctl/ovs-dpctl/
406 ovs-ofctl/ovsdb-tool commands. Please check
407 utilities/ovs-command-compgen.INSTALL.md for how to use.
408 - The "learn" action supports a new flag "delete_learned" that causes
409 the learned flows to be deleted when the flow with the "learn" action
411 - Basic support for the Geneve tunneling protocol. It is not yet
412 possible to generate or match options. This is planned for a future
413 release. The protocol is documented at
414 http://tools.ietf.org/html/draft-gross-geneve-00
415 - The OVS database now reports controller rate limiting statistics.
416 - sflow now exports information about LACP-based bonds, port names, and
417 OpenFlow port numbers, as well as datapath performance counters.
418 - ovs-dpctl functionality is now available for datapaths integrated
419 into ovs-vswitchd, via ovs-appctl. Some existing ovs-appctl
420 commands are now redundant and will be removed in a future
421 release. See ovs-vswitchd(8) for details.
423 * OpenFlow 1.4 bundles are now supported for flow mods and port
424 mods. For flow mods, both 'atomic' and 'ordered' bundle flags
425 are trivially supported, as all bundled messages are executed
426 in the order they were added and all flow table modifications
427 are now atomic to the datapath. Port mods may not appear in
428 atomic bundles, as port status modifications are not atomic.
429 * IPv6 flow label and neighbor discovery fields are now modifiable.
430 * OpenFlow 1.5 extended registers are now supported.
431 * The OpenFlow 1.5 actset_output field is now supported.
432 * OpenFlow 1.5 Copy-Field action is now supported.
433 * OpenFlow 1.5 masked Set-Field action is now supported.
434 * OpenFlow 1.3+ table features requests are now supported (read-only).
435 * Nicira extension "move" actions may now be included in action sets.
436 * "resubmit" actions may now be included in action sets. The resubmit
437 is executed last, and only if the action set has no "output" or "group"
439 * OpenFlow 1.4+ flow "importance" is now maintained in the flow table.
440 * A new Netronome extension to OpenFlow 1.5+ allows control over the
441 fields hashed for OpenFlow select groups. See "selection_method" and
442 related options in ovs-ofctl(8) for details.
443 - ovs-ofctl has a new '--bundle' option that makes the flow mod commands
444 ('add-flow', 'add-flows', 'mod-flows', 'del-flows', and 'replace-flows')
445 use an OpenFlow 1.4 bundle to operate the modifications as a single
446 atomic transaction. If any of the flow mods in a transaction fail, none
447 of them are executed. All flow mods in a bundle appear to datapath
448 lookups simultaneously.
449 - ovs-ofctl 'add-flow' and 'add-flows' commands now accept arbitrary flow
450 mods as an input by allowing the flow specification to start with an
451 explicit 'add', 'modify', 'modify_strict', 'delete', or 'delete_strict'
452 keyword. A missing keyword is treated as 'add', so this is fully
453 backwards compatible. With the new '--bundle' option all the flow mods
454 are executed as a single atomic transaction using an OpenFlow 1.4 bundle.
455 - ovs-pki: Changed message digest algorithm from MD5 to SHA-1 because
456 MD5 is no longer secure and some operating systems have started to disable
458 - ovsdb-server: New OVSDB protocol extension allows inequality tests on
459 "optional scalar" columns. See ovsdb-server(1) for details.
460 - ovs-vsctl now permits immutable columns in a new row to be modified in
461 the same transaction that creates the row.
462 - test-controller has been renamed ovs-testcontroller at request of users
463 who find it useful for testing basic OpenFlow setups. It is still not
464 a necessary or desirable part of most Open vSwitch deployments.
465 - Support for travis-ci.org based continuous integration builds has been
466 added. Build failures are reported to build@openvswitch.org. See INSTALL.md
467 file for additional details.
468 - Support for the Rapid Spanning Tree Protocol (IEEE 802.1D-2004).
469 The implementation has been tested successfully against the Ixia Automated
470 Network Validation Library (ANVL).
471 - Stats are no longer updated on fake bond interface.
472 - Keep active bond slave selection across OVS restart.
473 - A simple wrapper script, 'ovs-docker', to integrate OVS with Docker
474 containers. If and when there is a native integration of Open vSwitch
475 with Docker, the wrapper script will be retired.
476 - Added support for DPDK Tunneling. VXLAN, GRE, and Geneve are supported
477 protocols. This is generic tunneling mechanism for userspace datapath.
478 - Support for multicast snooping (IGMPv1, IGMPv2 and IGMPv3)
479 - Support for Linux kernels up to 4.0.x
480 - The documentation now use the term 'destination' to mean one of syslog,
481 console or file for vlog logging instead of the previously used term
483 - Support for VXLAN Group Policy extension
484 - Initial support for the IETF Auto-Attach SPBM draft standard. This
485 contains rudimentary support for the LLDP protocol as needed for
487 - The default OpenFlow and OVSDB ports are now the IANA-assigned
488 numbers. OpenFlow is 6653 and OVSDB is 6640.
489 - Support for DPDK vHost.
490 - Support for outer UDP checksums in Geneve and VXLAN.
491 - The kernel vports with dependencies are no longer part of the overall
492 openvswitch.ko but built and loaded automatically as individual kernel
493 modules (vport-*.ko).
494 - Support for STT tunneling.
495 - ovs-sim: New developer tool for simulating multiple OVS instances.
496 See ovs-sim(1) for more information.
497 - Support to configure method (--syslog-method argument) that determines
498 how daemons will talk with syslog.
499 - Support for "ovs-appctl vlog/list-pattern" command that lets to query
500 logging message format for each destination.
504 ---------------------
505 - OpenFlow 1.1, 1.2, and 1.3 are now enabled by default in
507 - Linux kernel datapath now has an exact match cache optimizing the
508 flow matching process.
509 - Datapath flows now have partially wildcarded tranport port field
510 matches. This reduces userspace upcalls, but increases the
511 number of different masks in the datapath. The kernel datapath
512 exact match cache removes the overhead of matching the incoming
513 packets with the larger number of masks, but when paired with an
514 older kernel module, some workloads may perform worse with the
516 - Compatibility with autoconf 2.63 (previously >=2.64)
518 v2.2.0 - Internal Release
519 ---------------------
520 - Internal ports are no longer brought up by default, because it
521 should be an administrator task to bring up devices as they are
523 - ovs-vsctl now reports when ovs-vswitchd fails to create a new port or
525 - Port creation and configuration errors are now stored in a new error
526 column of the Interface table and included in 'ovs-vsctl show'.
527 - The "ovsdbmonitor" graphical tool has been removed, because it was
528 poorly maintained and not widely used.
529 - New "check-ryu" Makefile target for running Ryu tests for OpenFlow
530 controllers against Open vSwitch. See INSTALL.md for details.
531 - Added IPFIX support for SCTP flows and templates for ICMPv4/v6 flows.
532 - Upon the receipt of a SIGHUP signal, ovs-vswitchd no longer reopens its
533 log file (it will terminate instead). Please use 'ovs-appctl vlog/reopen'
535 - Support for Linux kernels up to 3.14. From Kernel 3.12 onwards OVS uses
536 tunnel API for GRE and VXLAN.
537 - Added DPDK support.
538 - Added support for custom vlog patterns in Python
542 ---------------------
543 - Address prefix tracking support for flow tables. New columns
544 "prefixes" in OVS-DB table "Flow_Table" controls which packet
545 header fields are used for address prefix tracking. Prefix
546 tracking allows the classifier to skip rules with longer than
547 necessary prefixes, resulting in better wildcarding for datapath
548 flows. Default configuration is to not use any fields for prefix
549 tracking. However, if any flow tables contain both exact matches
550 and masked matches for IP address fields, OVS performance may be
551 increased by using this feature.
552 * As of now, the fields for which prefix lookup can be enabled
553 are: 'tun_id', 'tun_src', 'tun_dst', 'nw_src', 'nw_dst' (or
554 aliases 'ip_src' and 'ip_dst'), 'ipv6_src', and 'ipv6_dst'.
555 (Using this feature for 'tun_id' would only make sense if the
556 tunnel IDs have prefix structure similar to IP addresses.)
557 * There is a maximum number of fields that can be enabled for any
558 one flow table. Currently this limit is 3.
560 $ ovs-vsctl set Bridge br0 flow_tables:0=@N1 -- \
561 --id=@N1 create Flow_Table name=table0
562 $ ovs-vsctl set Bridge br0 flow_tables:1=@N1 -- \
563 --id=@N1 create Flow_Table name=table1
564 $ ovs-vsctl set Flow_Table table0 prefixes=ip_dst,ip_src
565 $ ovs-vsctl set Flow_Table table1 prefixes=[]
566 - TCP flags matching: OVS now supports matching of TCP flags. This
567 has an adverse performance impact when using OVS userspace 1.10
568 or older (no megaflows support) together with the new OVS kernel
569 module. It is recommended that the kernel and userspace modules
570 both are upgraded at the same time.
571 - The default OpenFlow and OVSDB ports will change to
572 IANA-assigned numbers in a future release. Consider updating
573 your installations to specify port numbers instead of using the
576 * The OpenFlow 1.1+ "Write-Actions" instruction is now supported.
577 * OVS limits the OpenFlow port numbers it assigns to port 32767 and
578 below, leaving port numbers above that range free for assignment
580 * ovs-vswitchd now honors changes to the "ofport_request" column
581 in the Interface table by changing the port's OpenFlow port
583 * The Open vSwitch software switch now supports OpenFlow groups.
584 - ovs-vswitchd.conf.db.5 man page will contain graphviz/dot
585 diagram only if graphviz package was installed at the build time.
586 - Support for Linux kernels up to 3.11
588 The "show" command also displays mega flow mask stats.
590 * New command "ofp-parse-pcap" to dump OpenFlow from PCAP files.
591 - ovs-controller has been renamed test-controller. It is no longer
592 packaged or installed by default, because too many users assumed
593 incorrectly that ovs-controller was a necessary or desirable part
594 of an Open vSwitch deployment.
595 - Added vlog option to export to a UDP syslog sink.
597 * The "monitor" command can now monitor all tables in a database,
598 instead of being limited to a single table.
599 - The flow-eviction-threshold has been replaced by the flow-limit which is a
600 hard limit on the number of flows in the datapath. It defaults to 200,000
601 flows. OVS automatically adjusts this number depending on network
603 - Added IPv6 support for active and passive socket communications.
607 ---------------------
608 - The ovs-vswitchd process is no longer single-threaded. Multiple
609 threads are now used to handle flow set up and asynchronous
612 * Experimental support for OpenFlow 1.1 (in addition to 1.2 and
613 1.3, which had experimental support in 1.10).
614 * Experimental protocol support for OpenFlow 1.1+ groups. This
615 does not yet include an implementation in the Open vSwitch
617 * Experimental protocol support for OpenFlow 1.2+ meters. This
618 does not yet include an implementation in the Open vSwitch
620 * New support for matching outer source and destination IP address
621 of tunneled packets, for tunnel ports configured with the newly
622 added "remote_ip=flow" and "local_ip=flow" options.
623 * Support for matching on metadata 'pkt_mark' for interacting with
624 other system components. On Linux this corresponds to the skb
626 * Support matching, rewriting SCTP ports
627 - The Interface table in the database has a new "ifindex" column to
628 report the interface's OS-assigned ifindex.
629 - New "check-oftest" Makefile target for running OFTest against Open
630 vSwitch. See README-OFTest for details.
631 - The flow eviction threshold has been moved to the Open_vSwitch table.
632 - Database names are now mandatory when specifying ovsdb-server options
633 through database paths (e.g. Private key option with the database name
634 should look like "--private-key=db:Open_vSwitch,SSL,private_key").
635 - Added ovs-dev.py, a utility script helpful for Open vSwitch developers.
636 - Support for Linux kernels up to 3.10
638 * New "ofp-parse" for printing OpenFlow messages read from a file.
639 * New commands for OpenFlow 1.1+ groups.
640 - Added configurable flow caching support to IPFIX exporter.
641 - Dropped support for Linux pre-2.6.32.
642 - Log file timestamps and ovsdb commit timestamps are now reported
643 with millisecond resolution. (Previous versions only reported
647 v1.11.0 - 28 Aug 2013
648 ---------------------
649 - Support for megaflows, which allows wildcarding in the kernel (and
650 any dpif implementation that supports wildcards). Depending on
651 the flow table and switch configuration, flow set up rates are
652 close to the Linux bridge.
653 - The "tutorial" directory contains a new tutorial for some advanced
654 Open vSwitch features.
655 - Stable bond mode has been removed.
656 - The autopath action has been removed.
657 - New support for the data encapsulation format of the LISP tunnel
658 protocol (RFC 6830). An external control plane or manual flow
659 setup is required for EID-to-RLOC mapping.
661 * The "dec_mpls_ttl" and "set_mpls_ttl" actions from OpenFlow
662 1.1 and later are now implemented.
663 * New "stack" extension for use in actions, to push and pop from
665 * The "load" and "set_field" actions can now modify the "in_port". (This
666 allows one to enable output to a flow's input port by setting the
667 in_port to some unused value, such as OFPP_NONE.)
669 * New debugging commands "add-flow", "mod-flow", "del-flow".
670 * "dump-flows" now has a -m option to increase output verbosity.
671 - In dpif-based bridges, cache action translations, which can improve
672 flow set up performance by 80% with a complicated flow table.
673 - New syslog format, prefixed with "ovs|", to be easier to filter.
674 - RHEL: Removes the default firewall rule that allowed GRE traffic to
675 pass through. Any users that relied on this automatic firewall hole
676 will have to manually configure it. The ovs-ctl(8) manpage documents
677 the "enable-protocol" command that can be used as an alternative.
678 - New CFM demand mode which uses data traffic to indicate interface
681 v1.10.0 - 01 May 2013
682 ---------------------
683 - Bridge compatibility support has been removed. Any uses that
684 rely on ovs-brcompatd will have to stick with Open vSwitch 1.9.x
685 or adapt to native Open vSwitch support (e.g. use ovs-vsctl instead
687 - The maximum size of the MAC learning table is now configurable.
688 - With the Linux datapath, packets for new flows are now queued
689 separately on a per-port basis, so it should no longer be
690 possible for a large number of new flows arriving on one port to
691 prevent new flows from being processed on other ports.
693 * Previously ovs-vsctl would retry connecting to the database forever,
694 causing it to hang if ovsdb-server was not running. Now, ovs-vsctl
695 only tries once by default (use --retry to try forever). This change
696 means that you may want to remove uses of --timeout to avoid hangs
698 * Many "ovs-vsctl" database commands now accept an --if-exists option.
699 Please refer to the ovs-vsctl manpage for details.
701 - Experimental support for newer versions of OpenFlow. See
702 the "What versions of OpenFlow does Open vSwitch support?"
703 question in the FAQ for more details.
704 - The OpenFlow "dp_desc" may now be configured by setting the
705 value of other-config:dp-desc in the Bridge table.
706 - It is possible to request the OpenFlow port number with the
707 "ofport_request" column in the Interface table.
708 - The NXM flow_removed message now reports the OpenFlow table ID
709 from which the flow was removed.
711 - New support for the VXLAN tunnel protocol (see the IETF draft here:
712 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03).
713 - Tunneling requires the version of the kernel module paired with
714 Open vSwitch 1.9.0 or later.
715 - Inheritance of the Don't Fragment bit in IP tunnels (df_inherit)
716 is no longer supported.
717 - Path MTU discovery is no longer supported.
718 - CAPWAP tunneling support removed.
719 - Tunnels with multicast destination ports are no longer supported.
721 - The "dump-flows" and "del-flows" no longer require an argument
722 if only one datapath exists.
724 - New "vlog/disable-rate-limit" and "vlog/enable-rate-limit"
725 commands available allow control over logging rate limits.
726 - New "dpif/dump-dps", "dpif/show", and "dpif/dump-flows" command
727 that mimic the equivalent ovs-dpctl commands.
728 - The ofproto library is now responsible for assigning OpenFlow port
729 numbers. An ofproto implementation should assign them when
730 port_construct() is called.
731 - All dpif-based bridges of a particular type share a common
732 datapath called "ovs-<type>", e.g. "ovs-system". The ovs-dpctl
733 commands will now return information on that shared datapath. To
734 get the equivalent bridge-specific information, use the new
735 "ovs-appctl dpif/*" commands.
736 - Backward-incompatible changes:
737 - Earlier Open vSwitch versions treated ANY as a wildcard in flow
738 syntax. OpenFlow 1.1 adds a port named ANY, which introduces a
739 conflict. ANY was rarely used in flow syntax, so we chose to
740 retire that meaning of ANY in favor of the OpenFlow 1.1 meaning.
741 - Patch ports no longer require kernel support, so they now work
742 with FreeBSD and the kernel module built into Linux 3.3 and later.
743 - New "sample" action.
747 ------------------------
749 - Support for ipv6 set action.
750 - SKB mark matching and setting.
751 - support for Linux kernels up to 3.8
752 - FreeBSD is now a supported platform, thanks to code contributions from
753 Gaetano Catalli, Ed Maste, and Giuseppe Lettieri.
754 - ovs-bugtool: New --ovs option to report only OVS related information.
755 - New %t and %T log escapes to identify the subprogram within a
756 cooperating group of processes or threads that emitted a log message.
757 The default log patterns now include this information.
759 - Allow bitwise masking for SHA and THA fields in ARP, SLL and TLL
760 fields in IPv6 neighbor discovery messages, and IPv6 flow label.
761 - Adds support for writing to the metadata field for a flow.
763 - The tunneling code no longer assumes input and output keys are
764 symmetric. If they are not, PMTUD needs to be disabled for
765 tunneling to work. Note this only applies to flow-based keys.
766 - New support for a nonstandard form of GRE that supports a 64-bit key.
767 - Tunnel Path MTU Discovery default value was set to 'disabled'.
768 This feature is deprecated and will be removed soon.
769 - Tunnel header caching removed.
771 - Commands and actions that accept port numbers now also accept keywords
772 that represent those ports (such as LOCAL, NONE, and ALL). This is
773 also the recommended way to specify these ports, for compatibility
774 with OpenFlow 1.1 and later (which use the OpenFlow 1.0 numbers
775 for these ports for different purposes).
777 - Support requesting the port number with the "port_no" option in
778 the "add-if" command.
779 - ovs-pki: The "online PKI" features have been removed, along with
780 the ovs-pki-cgi program that facilitated it, because of some
781 alarmist insecurity claims. We do not believe that these claims
782 are true, but because we do not know of any users for this
783 feature it seems better on balance to remove it. (The ovs-pki-cgi
784 program was not included in distribution packaging.)
785 - ovsdb-server now enforces the immutability of immutable columns. This
786 was not enforced in earlier versions due to an oversight.
787 - The following features are now deprecated. They will be removed no
788 earlier than February 2013. Please email dev@openvswitch.org with
790 - Bridge compatibility.
792 - The autopath action.
793 - Interface type "null".
794 - Numeric values for reserved ports (see "ovs-ofctl" note above).
795 - Tunnel Path MTU Discovery.
796 - CAPWAP tunnel support.
797 - The data in the RARP packets can now be matched in the same way as the
802 ------------------------
803 *** Internal only release ***
804 - New FAQ. Please send updates and additions!
805 - Authors of controllers, please read the new section titled "Action
806 Reproduction" in DESIGN, which describes an Open vSwitch change in
807 behavior in corner cases that may affect some controllers.
809 - A new test utility that can create L3 tunnel between two Open
810 vSwitches and detect connectivity issues.
812 - New --sort and --rsort options for "dump-flows" command.
813 - "mod-port" command can now control all OpenFlow config flags.
815 - Allow general bitwise masking for IPv4 and IPv6 addresses in
816 IPv4, IPv6, and ARP packets. (Previously, only CIDR masks
818 - Allow support for arbitrary Ethernet masks. (Previously, only
819 the multicast bit in the destination address could be individually
821 - New field OXM_OF_METADATA, to align with OpenFlow 1.1.
822 - The OFPST_QUEUE request now reports an error if a specified port or
823 queue does not exist, or for requests for a specific queue on all
824 ports, if the specified queue does not exist on any port. (Previous
825 versions generally reported an empty set of results.)
826 - New "flow monitor" feature to allow controllers to be notified of
827 flow table changes as they happen.
828 - Additional protocols are not mirrored and dropped when forward-bpdu is
829 false. For a full list, see the ovs-vswitchd.conf.db man page.
830 - Open vSwitch now sends RARP packets in situations where it previously
831 sent a custom protocol, making it consistent with behavior of QEMU and
833 - All Open vSwitch programs and log files now show timestamps in UTC,
834 instead the local timezone, by default.
838 ------------------------
839 - kernel modules are renamed. openvswitch_mod.ko is now
840 openvswitch.ko and brcompat_mod.ko is now brcompat.ko.
841 - Increased the number of NXM registers to 8.
842 - Added ability to configure DSCP setting for manager and controller
843 connections. By default, these connections have a DSCP value of
844 Internetwork Control (0xc0).
845 - Added the granular link health statistics, 'cfm_health', to an
848 - Added support to mask nd_target for ICMPv6 neighbor discovery flows.
849 - Added support for OpenFlow 1.3 port description (OFPMP_PORT_DESC)
852 - Added the "dump-ports-desc" command to retrieve port
853 information using the new port description multipart messages.
855 - Added support for spawning ovs-test server from the client.
856 - Now ovs-test is able to automatically create test bridges and ports.
857 - "ovs-dpctl dump-flows" now prints observed TCP flags in TCP flows.
858 - Tripled flow setup performance.
859 - The "coverage/log" command previously available through ovs-appctl
860 has been replaced by "coverage/show". The new command replies with
861 coverage counter values, instead of logging them.
865 ------------------------
866 - Allow OFPP_CONTROLLER as the in_port for packet-out messages.
870 ------------------------
871 *** Internal only release ***
873 - LACP bonds no longer fall back to balance-slb when negotiations fail.
874 Instead they drop traffic.
875 - The default bond_mode changed from SLB to active-backup, to protect
876 unsuspecting users from the significant risks of SLB bonds (which are
877 documented in vswitchd/INTERNALS).
878 - Load balancing can be disabled by setting the bond-rebalance-interval
881 - Added support for bitwise matching on TCP and UDP ports.
882 See ovs-ofctl(8) for more information.
883 - NXM flow dumps now include times elapsed toward idle and hard
885 - Added an OpenFlow extension NXT_SET_ASYNC_CONFIG that allows
886 controllers more precise control over which OpenFlow messages they
887 receive asynchronously.
888 - New "fin_timeout" action.
889 - Added "fin_timeout" support to "learn" action.
890 - New Nicira action NXAST_CONTROLLER that offers additional features
891 over output to OFPP_CONTROLLER.
892 - When QoS settings for an interface do not configure a default queue
893 (queue 0), Open vSwitch now uses a default configuration for that
894 queue, instead of dropping all packets as in previous versions.
896 - Logging to console and file will have UTC timestamp as a default for
897 all the daemons. An example of the default format is
898 2012-01-27T16:35:17Z. ovs-appctl can be used to change the default
900 - The syntax of commands and options to set log levels was simplified,
901 to make it easier to remember.
902 - New support for limiting the number of flows in an OpenFlow flow
903 table, with configurable policy for evicting flows upon
904 overflow. See the Flow_Table table in ovs-vswitch.conf.db(5)
905 for more information.
906 - New "enable-async-messages" column in the Controller table. If set to
907 false, OpenFlow connections to the controller will initially have all
908 asynchronous messages disabled, overriding normal OpenFlow behavior.
909 - ofproto-provider interface:
910 - "struct rule" has a new member "used" that ofproto implementations
911 should maintain by updating with ofproto_rule_update_used().
913 - The new option --timestamp causes the "monitor" command to print
914 a timestamp with every update.
915 - CFM module CCM broadcasts can now be tagged with an 802.1p priority.
919 ------------------------
921 - Added support for querying, modifying, and deleting flows
922 based on flow cookie when using NXM.
923 - Added new NXM_PACKET_IN format.
924 - Added new NXAST_DEC_TTL action.
926 - Added daemonization support to the monitor and snoop commands.
928 - The "find" command supports new set relational operators
929 {=}, {!=}, {<}, {>}, {<=}, and {>=}.
930 - ovsdb-tool now uses the typical database and schema installation
931 directories as defaults.
932 - The default MAC learning timeout has been increased from 60 seconds
933 to 300 seconds. The MAC learning timeout is now configurable.
937 ------------------------
938 - Compatible with Open vSwitch kernel module included in Linux 3.3.
939 - New "VLAN splinters" feature to work around buggy device drivers
940 in old Linux versions. (This feature is deprecated. When
941 broken device drivers are no longer in widespread use, we will
942 delete this feature.) See ovs-vswitchd.conf.db(5) for more
945 - Added ability to match on IPv6 flow label through NXM.
946 - Added ability to match on ECN bits in IPv4 and IPv6 through NXM.
947 - Added ability to match on TTL in IPv4 and IPv6 through NXM.
948 - Added ability to modify ECN bits in IPv4.
949 - Added ability to modify TTL in IPv4.
951 - Don't require the "normal" action to use mirrors. Traffic will
952 now be properly mirrored for any flows, regardless of their
954 - Track packet and byte statistics sent on mirrors.
955 - The sFlow implementation can now usually infer the correct agent
956 device instead of having to be told explicitly.
958 - New "fdb/flush" command to flush bridge's MAC learning table.
960 - A new distributed testing tool that allows one to diagnose performance
961 and connectivity issues. This tool currently is not included in RH or
963 - RHEL packaging now supports integration with Red Hat network scripts.
965 - Post 1.4.*, OVS will be changing the default bond mode from balance-slb
966 to active-backup. SLB bonds carry significant risks with them
967 (documented vswitchd/INTERNALS) which we want to prevent unsuspecting
968 users from running into. Users are advised to update any scripts or
969 configuration which may be negatively impacted by explicitly setting
970 the bond mode which they want to use.
974 ------------------------
976 - Added an OpenFlow extension which allows the "output" action to accept
978 - Added an OpenFlow extension for flexible learning.
979 - Bumped number of NXM registers from four to five.
981 - New "version" command to determine version of running daemon.
982 - If no argument is provided for "cfm/show", displays detailed
983 information about all interfaces with CFM enabled.
984 - If no argument is provided for "lacp/show", displays detailed
985 information about all ports with LACP enabled.
987 - New "set-if" command to modify a datapath port's configuration.
989 - The software switch now supports 255 OpenFlow tables, instead
990 of just one. By default, only table 0 is consulted, but the
991 new NXAST_RESUBMIT_TABLE action can look up in additional
992 tables. Tables 128 and above are reserved for use by the
993 switch itself; please use only tables 0 through 127.
994 - Add support for 802.1D spanning tree (STP).
995 - Fragment handling extensions:
996 - New OFPC_FRAG_NX_MATCH fragment handling mode, in which L4
997 fields are made available for matching in fragments with
999 - New NXM_NX_IP_FRAG match field for matching IP fragments (usable
1000 via "ip_frag" in ovs-ofctl).
1001 - New ovs-ofctl "get-frags" and "set-frags" commands to get and set
1002 fragment handling policy.
1003 - CAPWAP tunneling now supports an extension to transport a 64-bit key.
1004 By default it remains compatible with the old version and other
1005 standards-based implementations.
1006 - Flow setups are now processed in a round-robin manner across ports
1007 to prevent any single client from monopolizing the CPU and conducting
1008 a denial of service attack.
1009 - Added support for native VLAN tagging. A new "vlan_mode"
1010 parameter can be set for "port". Possible values: "access",
1011 "trunk", "native-tagged" and "native-untagged".
1012 - test-openflowd has been removed. Please use ovs-vswitchd instead.
1014 v1.2.0 - 03 Aug 2011
1015 ------------------------
1016 - New "ofproto" abstraction layer to ease porting to hardware
1018 - Packaging for Red Hat Enterprise Linux 5.6 and 6.0.
1019 - Datapath support for Linux kernels up to 3.0.
1021 - New "bundle" and "bundle_load" action extensions.
1023 - Implement table unique constraints.
1024 - Support cooperative locking between callers.
1026 - New "-s" option for "show" command prints packet and byte
1027 counters for each port.
1029 - New "--readd" option for "replace-flows".
1031 - New "show" command to print an overview of configuration.
1032 - New "comment" command to add remark that explains intentions.
1033 - ovs-brcompatd has been rewritten to fix long-standing bugs.
1034 - ovs-openflowd has been renamed test-openflowd and moved into the
1035 tests directory. Its presence confused too many users. Please
1036 use ovs-vswitchd instead.
1037 - New ovs-benchmark utility to test flow setup performance.
1038 - A new log level "off" has been added. Configuring a log facility
1039 "off" prevents any messages from being logged to it. Previously,
1040 "emer" was effectively "off" because no messages were ever logged at
1041 level "emer". Now, errors that cause a process to exit are logged
1043 - "configure" option --with-l26 has been renamed --with-linux, and
1044 --with-l26-source has been renamed --with-linux-source. The old
1045 names will be removed after the next release, so please update
1047 - The "-2.6" suffix has been dropped from the datapath/linux-2.6 and
1048 datapath/linux-2.6/compat-2.6 directories.
1050 - Dropped support for "tun_id_from_cookie" OpenFlow extension.
1051 Please use the extensible match extensions instead.
1052 - Removed the Maintenance_Point and Monitor tables in an effort
1053 to simplify 802.1ag configuration.
1054 - Performance and scalability improvements
1057 v1.1.0 - 05 Apr 2011
1058 ------------------------
1059 - Ability to define policies over IPv6
1062 - Support for extensible match extensions to OpenFlow
1064 - Support for HFSC qdisc.
1065 - Queue used by in-band control can now be configured.
1067 - Kernel<->userspace interface has been reworked and should be
1068 close to a stable ABI now.
1069 - "Port group" concept has been dropped.
1070 - GRE over IPSEC tunnels
1072 - New active backup bonding mode.
1073 - New L4 hashing support when LACP is enabled.
1074 - Source MAC hash now includes VLAN field also.
1076 - Greatly improved handling of large flow tables
1078 - "show" command now prints full vport configuration.
1079 - "dump-groups" command removed since kernel support for
1080 port groups was dropped.
1082 - New commands for working with the new Managers table.
1083 - "list" command enhanced with new formatting options and --columns
1085 - "get" command now accepts new --id option.
1086 - New "find" command.
1088 - New "queue-stats" command for printing queue stats.
1089 - New commands "replace-flows" and "diff-flows".
1090 - Commands to add and remove flows can now read from files.
1091 - New --flow-format option to enable or disable NXM.
1092 - New --more option to increase OpenFlow message verbosity.
1093 - Removed "tun-cookie" command, which is no longer useful.
1094 - ovs-controller enhancements for testing various features.
1095 - New ovs-vlan-test command for testing for Linux kernel driver VLAN
1096 bugs. New ovs-vlan-bug-workaround command for enabling and
1097 disabling a workaround for these driver bugs.
1099 - "Resubmit" actions now update flow statistics.
1100 - New "register" extension for use in matching and actions, via NXM.
1101 - New "multipath" experimental action extension.
1102 - New support for matching multicast Ethernet frames, via NXM.
1103 - New extension for OpenFlow vendor error codes.
1104 - New extension to set the QoS output queue without actually
1105 sending to an output port.
1106 - Open vSwitch now reports a single flow table, instead of
1107 separate hash and wildcard tables. This better models the
1108 current implementation.
1109 - New experimental "note" action.
1110 - New "ofproto/trace" ovs-appctl command and associated utilities
1111 to ease debugging complex flow tables.
1113 - Schema documentation now includes an entity-relationship diagram.
1114 - The database is now garbage collected. In most tables,
1115 unreferenced rows will be deleted automatically.
1116 - Many tables now include statistics updated periodically by
1117 ovs-vswitchd or ovsdb-server.
1118 - Every table now has an "external-ids" column for use by OVS
1120 - There is no default controller anymore. Each bridge must have its
1121 controller individually specified.
1122 - The "fail-mode" is now a property of a Bridge instead of a Controller.
1123 - New versioning and checksum features.
1124 - New Managers table and manager_options column in Open_vSwitch table
1125 for specifying managers. The old "managers" column in the
1126 Open_vSwitch table has been removed.
1127 - Many "name" columns are now immutable.
1129 - Dropped support for XenServer pre-5.6.100.
1130 - Dropped support for Linux pre-2.6.18.
1131 - Dropped controller discovery support.
1132 - Dropped "ovs-ofctl status" and the OpenFlow extension that it used.
1133 Statistics reporting in the database is a rough equivalent.
1134 - Dropped the "corekeeper" package (now separate, at
1135 http://openvswitch.org/cgi-bin/gitweb.cgi?p=corekeeper).
1136 - Performance and scalability improvements
1139 v1.1.0pre2 - 13 Sep 2010
1140 ------------------------
1143 v1.1.0pre1 - 31 Aug 2010
1144 ------------------------
1145 - OpenFlow 1.0 slicing (QoS) functionality
1146 - Python bindings for configuration database (no write support)
1147 - Performance and scalability improvements
1150 v1.0.1 - 31 May 2010
1151 --------------------
1152 - New "patch" interface type
1155 v1.0.0 - 15 May 2010
1156 --------------------
1157 - Configuration database with remote management
1160 - Support for XenServer 5.5 and 5.6
1161 - Performance and scalability improvements
1164 v0.99.2 - 18 Feb 2010
1165 ---------------------
1168 v0.99.1 - 25 Jan 2010
1169 ---------------------
1170 - Add support for sFlow(R)
1171 - Make headers compatible with C++
1174 v0.99.0 - 14 Jan 2010
1175 ---------------------
1176 - User-space forwarding engine
1179 v0.90.7 - 29 Nov 2009
1180 ---------------------
1181 - Add support for NetFlow active timeouts
1184 v0.90.6 - 6 Oct 2009
1185 --------------------
1188 v0.90.5 - 21 Sep 2009
1189 ---------------------
1190 - Generalize in-band control to more diverse network setups
projects / ovs.git / blob