3 - Fixed regression in table stats maintenance introduced in OVS
4 2.3.0, wherein the number of OpenFlow table hits and misses was
7 * OFPT_PACKET_OUT messages are now supported in bundles.
8 * A new "selection_method=dp_hash" type for OpenFlow select group
9 bucket selection that uses the datapath computed 5-tuple hash
10 without making datapath flows match the 5-tuple fields, which
11 is useful for more efficient load balancing, for example. This
12 uses the Netronome extension to OpenFlow 1.5+ that allows
13 control over the OpenFlow select groups selection method. See
14 "selection_method" and related options in ovs-ofctl(8) for
17 * 'bundle' command now supports packet-out messages.
18 * New syntax for 'ovs-ofctl packet-out' command, which uses the
19 same string parser as the 'bundle' command. The old 'packet-out'
20 syntax is deprecated and will be removed in a later OVS
22 * New unixctl "ofctl/packet-out" command, which can be used to
23 instruct a flow monitor to issue OpenFlow packet-out messages.
25 * TLV mappings for protocols such as Geneve are now segregated on
26 a per-OpenFlow bridge basis rather than globally. (The interface
31 - First supported release of OVN. See ovn-architecture(7) for more
34 * New "monitor_cond" "monitor_cond_update" and "update2" extensions to
37 * OpenFlow 1.3+ bundles now expire after 10 seconds since the
38 last time the bundle was either opened, modified, or closed.
39 * OpenFlow 1.3 Extension 230, adding OpenFlow Bundles support, is
41 * OpenFlow 1.3+ bundles are now supported for group mods as well as
42 flow mods and port mods. Both 'atomic' and 'ordered' bundle
43 flags are supported for group mods as well as flow mods.
44 * Internal OpenFlow rule representation for load and set-field
45 actions is now much more memory efficient. For a complex flow
46 table this can reduce rule memory consumption by 40%.
47 * Bundles are now much more memory efficient than in OVS 2.5.
48 Together with memory efficiency improvements in OpenFlow rule
49 representation, the peak OVS resident memory use during a
50 bundle commit for large complex set of flow mods can be only
51 25% of that in OVS 2.5 (4x lower).
52 * OpenFlow 1.1+ OFPT_QUEUE_GET_CONFIG_REQUEST now supports OFPP_ANY.
53 * OpenFlow 1.4+ OFPMP_QUEUE_DESC is now supported.
54 * OpenFlow 1.4+ OFPT_TABLE_STATUS is now supported.
55 * New property-based packet-in message format NXT_PACKET_IN2 with support
56 for arbitrary user-provided data and for serializing flow table
57 traversal into a continuation for later resumption.
58 * New extension message NXT_SET_ASYNC_CONFIG2 to allow OpenFlow 1.4-like
59 control over asynchronous messages in earlier versions of OpenFlow.
60 * New OpenFlow extension NXM_NX_MPLS_TTL to provide access to MPLS TTL.
61 * New output option, output(port=N,max_len=M), to allow truncating a
62 packet to size M bytes when outputting to port N.
63 * New command OFPGC_ADD_OR_MOD for OFPT_GROUP_MOD message that adds a
64 new group or modifies an existing groups
65 * The optional OpenFlow packet buffering feature is deprecated in
66 this release, and will be removed in the next OVS release
67 (2.7). After the change OVS always sends the 'buffer_id' as
68 0xffffffff in packet-in messages and will send an error
69 response if any other value of this field is included in
70 packet-out and flow mod sent by a controller. Controllers are
71 already expected to work properly in cases where the switch can
72 not buffer packets, so this change should not affect existing
74 * New OpenFlow extension NXT_CT_FLUSH_ZONE to flush conntrack zones.
75 - Improved OpenFlow version compatibility for actions:
76 * New OpenFlow extension to support the "group" action in OpenFlow 1.0.
77 * OpenFlow 1.0 "enqueue" action now properly translated to OpenFlow 1.1+.
78 * OpenFlow 1.1 "mod_nw_ecn" and OpenFlow 1.1+ "mod_nw_ttl" actions now
79 properly translated to OpenFlow 1.0.
81 * queue-get-config command now allows a queue ID to be specified.
82 * '--bundle' option can now be used with OpenFlow 1.3 and with group mods.
83 * New "bundle" command allows executing a mixture of flow and group mods
84 as a single atomic transaction.
85 * New option "--color" to produce colorized output for some commands.
86 * New option '--may-create' to use OFPGC_ADD_OR_MOD in mod-group command.
88 * New "sampling_port" option for "sample" action to allow sampling
89 ingress and egress tunnel metadata with IPFIX.
90 * New ovs-ofctl commands "dump-ipfix-bridge" and "dump-ipfix-flow" to
91 dump bridge IPFIX statistics and flow based IPFIX statistics.
92 * New setting other-config:virtual_obs_id to add an arbitrary string
95 * OVS Linux datapath now implements Conntrack NAT action with all
96 supported Linux kernels.
97 * Support for truncate action.
98 * New QoS type "linux-noop" that prevents Open vSwitch from trying to
99 manage QoS for a given port (useful when other software manages QoS).
101 * New option "n_rxq" for PMD interfaces.
102 Old 'other_config:n-dpdk-rxqs' is no longer supported.
103 Not supported by vHost interfaces. For them number of rx and tx queues
104 is applied from connected virtio device.
105 * New 'other_config:pmd-rxq-affinity' field for PMD interfaces, that
106 allows to pin port's rx queues to desired cores.
107 * New appctl command 'dpif-netdev/pmd-rxq-show' to check the port/rxq
109 * Type of log messages from PMD threads changed from INFO to DBG.
110 * QoS functionality with sample egress-policer implementation.
111 * The mechanism for configuring DPDK has changed to use database
112 * Sensible defaults have been introduced for many of the required
113 configuration options
114 * DB entries have been added for many of the DPDK EAL command line
115 arguments. Additional arguments can be passed via the dpdk-extra
117 * Add ingress policing functionality.
118 * PMD threads servicing vHost User ports can now come from the NUMA
119 node that device memory is located on if CONFIG_RTE_LIBRTE_VHOST_NUMA
121 * Basic connection tracking for the userspace datapath (no ALG,
122 fragmentation or NAT support yet)
123 * Support for DPDK 16.07
124 * Optional support for DPDK pdump enabled.
125 * Jumbo frame support
126 * Remove dpdkvhostcuse port type.
127 * OVS client mode for vHost and vHost reconnect (Requires QEMU 2.7)
128 * 'dpdkvhostuserclient' port type.
129 - Increase number of registers to 16.
130 - ovs-benchmark: This utility has been removed due to lack of use and
133 * New "vlog/close" command.
135 * Added the ability to selectively start the forwarding and database
136 functions (ovs-vswitchd and ovsdb-server, respectively).
138 * Remove max number of sessions limit, to enable connection scaling
141 * Added support for Python 3.4+ in addition to existing support
144 * Introduced SELinux policy package.
145 - Datapath Linux kernel compatibility.
146 * Dropped support for kernel older than 3.10.
147 * Removed VLAN splinters feature.
148 * Datapath supports kernel upto 4.7.
150 * Flow based tunnel match and action can be used for IPv6 address using
151 tun_ipv6_src, tun_ipv6_dst fields.
152 * Added support for IPv6 tunnels, for details checkout FAQ.
153 * Deprecated support for IPsec tunnels ports.
154 - A wrapper script, 'ovs-tcpdump', to easily port-mirror an OVS port and
156 - Introduce --no-self-confinement flag that allows daemons to work with
157 sockets outside their run directory.
158 - ovs-pki: Changed message digest algorithm from SHA-1 to SHA-512 because
159 SHA-1 is no longer secure and some operating systems have started to
160 disable it in OpenSSL.
161 - Add 'mtu_request' column to the Interface table. It can be used to
162 configure the MTU of the ports.
165 - Using openvswitch module in conjunction with upstream Linux tunnels:
166 * When using the openvswitch module distributed with OVS against kernel
167 versions 4.4 to 4.6, the openvswitch module cannot be loaded or used at
168 the same time as "ip_gre".
169 - Conntrack FTP ALGs: When using the openvswitch module distributed with
170 OVS, particular Linux distribution kernels versions may provide diminished
171 functionality. This typically affects active FTP data connections when
172 using "actions=ct(alg=ftp),..." in flow tables. Specifically:
173 * Centos 7.1 kernels (3.10.0-2xx) kernels are unable to correctly set
174 up expectations for FTP data connections in multiple zones,
175 eg "actions=ct(zone=1,alg=ftp),ct(zone=2,alg=ftp),...". Executing the
176 "ct" action for subsequent data connections may fail to determine that
177 the data connection is "related" to an existing connection.
178 * Centos 7.2 kernels (3.10.0-3xx) kernels may not establish FTP ALG state
179 correctly for NATed connections. As a result, flows that perform NAT,
180 eg "actions=ct(nat,ftp=alg,table=1),..." may fail to NAT the packet,
181 and will populate the "ct_state=inv" bit in the flow.
185 ---------------------
186 - Dropped support for Python older than version 2.7. As a consequence,
187 using Open vSwitch 2.5 or later on XenServer 6.5 or earlier (which
188 have Python 2.4) requires first installing Python 2.7.
190 * Group chaining (where one OpenFlow group triggers another) is
192 * OpenFlow 1.4+ "importance" is now considered for flow eviction.
193 * OpenFlow 1.4+ OFPTC_EVICTION is now implemented.
194 * OpenFlow 1.4+ OFPTC_VACANCY_EVENTS is now implemented.
195 * OpenFlow 1.4+ OFPMP_TABLE_DESC is now implemented.
196 * Allow modifying the ICMPv4/ICMPv6 type and code fields.
197 * OpenFlow 1.4+ OFPT_SET_ASYNC_CONFIG and OFPT_GET_ASYNC_CONFIG are
200 * New "out_group" keyword for OpenFlow 1.1+ matching on output group.
202 * Geneve tunnels can now match and set options and the OAM bit.
203 * The nonstandard GRE64 tunnel extension has been dropped.
204 - Support Multicast Listener Discovery (MLDv1 and MLDv2).
205 - Add 'symmetric_l3l4' and 'symmetric_l3l4+udp' hash functions.
206 - sFlow agent now reports tunnel and MPLS structures.
207 - New 'check-system-userspace', 'check-kmod' and 'check-kernel' Makefile
208 targets to run a new system testsuite. These tests can be run inside
209 a Vagrant box. See INSTALL.md for details
210 - Mark --syslog-target argument as deprecated. It will be removed in
211 the next OVS release.
212 - Added --user option to all daemons
213 - Add support for connection tracking through the new "ct" action
214 and "ct_state"/"ct_zone"/"ct_mark"/"ct_label" match fields. Only
215 available on Linux kernels with the connection tracking module loaded.
216 - Add experimental version of OVN. OVN, the Open Virtual Network, is a
217 system to support virtual network abstraction. OVN complements the
218 existing capabilities of OVS to add native support for virtual network
219 abstractions, such as virtual L2 and L3 overlays and security groups.
221 * DPDK ports may now be created via network scripts (see README.RHEL).
224 * Added multiqueue support to vhost-user
225 * Note: QEMU 2.5+ required for multiqueue support
228 ---------------------
229 - Flow table modifications are now atomic, meaning that each packet
230 now sees a coherent version of the OpenFlow pipeline. For
231 example, if a controller removes all flows with a single OpenFlow
232 "flow_mod", no packet sees an intermediate version of the OpenFlow
233 pipeline where only some of the flows have been deleted.
234 - Added support for SFQ, FQ_CoDel and CoDel qdiscs.
235 - Add bash command-line completion support for ovs-vsctl Please check
236 utilities/ovs-command-compgen.INSTALL.md for how to use.
237 - The MAC learning feature now includes per-port fairness to mitigate
238 MAC flooding attacks.
239 - New support for a "conjunctive match" OpenFlow extension, which
240 allows constructing OpenFlow matches of the form "field1 in
241 {a,b,c...} AND field2 in {d,e,f...}" and generalizations. For details,
242 see documentation for the "conjunction" action in ovs-ofctl(8).
243 - Add bash command-line completion support for ovs-appctl/ovs-dpctl/
244 ovs-ofctl/ovsdb-tool commands. Please check
245 utilities/ovs-command-compgen.INSTALL.md for how to use.
246 - The "learn" action supports a new flag "delete_learned" that causes
247 the learned flows to be deleted when the flow with the "learn" action
249 - Basic support for the Geneve tunneling protocol. It is not yet
250 possible to generate or match options. This is planned for a future
251 release. The protocol is documented at
252 http://tools.ietf.org/html/draft-gross-geneve-00
253 - The OVS database now reports controller rate limiting statistics.
254 - sflow now exports information about LACP-based bonds, port names, and
255 OpenFlow port numbers, as well as datapath performance counters.
256 - ovs-dpctl functionality is now available for datapaths integrated
257 into ovs-vswitchd, via ovs-appctl. Some existing ovs-appctl
258 commands are now redundant and will be removed in a future
259 release. See ovs-vswitchd(8) for details.
261 * OpenFlow 1.4 bundles are now supported for flow mods and port
262 mods. For flow mods, both 'atomic' and 'ordered' bundle flags
263 are trivially supported, as all bundled messages are executed
264 in the order they were added and all flow table modifications
265 are now atomic to the datapath. Port mods may not appear in
266 atomic bundles, as port status modifications are not atomic.
267 * IPv6 flow label and neighbor discovery fields are now modifiable.
268 * OpenFlow 1.5 extended registers are now supported.
269 * The OpenFlow 1.5 actset_output field is now supported.
270 * OpenFlow 1.5 Copy-Field action is now supported.
271 * OpenFlow 1.5 masked Set-Field action is now supported.
272 * OpenFlow 1.3+ table features requests are now supported (read-only).
273 * Nicira extension "move" actions may now be included in action sets.
274 * "resubmit" actions may now be included in action sets. The resubmit
275 is executed last, and only if the action set has no "output" or "group"
277 * OpenFlow 1.4+ flow "importance" is now maintained in the flow table.
278 * A new Netronome extension to OpenFlow 1.5+ allows control over the
279 fields hashed for OpenFlow select groups. See "selection_method" and
280 related options in ovs-ofctl(8) for details.
281 - ovs-ofctl has a new '--bundle' option that makes the flow mod commands
282 ('add-flow', 'add-flows', 'mod-flows', 'del-flows', and 'replace-flows')
283 use an OpenFlow 1.4 bundle to operate the modifications as a single
284 atomic transaction. If any of the flow mods in a transaction fail, none
285 of them are executed. All flow mods in a bundle appear to datapath
286 lookups simultaneously.
287 - ovs-ofctl 'add-flow' and 'add-flows' commands now accept arbitrary flow
288 mods as an input by allowing the flow specification to start with an
289 explicit 'add', 'modify', 'modify_strict', 'delete', or 'delete_strict'
290 keyword. A missing keyword is treated as 'add', so this is fully
291 backwards compatible. With the new '--bundle' option all the flow mods
292 are executed as a single atomic transaction using an OpenFlow 1.4 bundle.
293 - ovs-pki: Changed message digest algorithm from MD5 to SHA-1 because
294 MD5 is no longer secure and some operating systems have started to disable
296 - ovsdb-server: New OVSDB protocol extension allows inequality tests on
297 "optional scalar" columns. See ovsdb-server(1) for details.
298 - ovs-vsctl now permits immutable columns in a new row to be modified in
299 the same transaction that creates the row.
300 - test-controller has been renamed ovs-testcontroller at request of users
301 who find it useful for testing basic OpenFlow setups. It is still not
302 a necessary or desirable part of most Open vSwitch deployments.
303 - Support for travis-ci.org based continuous integration builds has been
304 added. Build failures are reported to build@openvswitch.org. See INSTALL.md
305 file for additional details.
306 - Support for the Rapid Spanning Tree Protocol (IEEE 802.1D-2004).
307 The implementation has been tested successfully against the Ixia Automated
308 Network Validation Library (ANVL).
309 - Stats are no longer updated on fake bond interface.
310 - Keep active bond slave selection across OVS restart.
311 - A simple wrapper script, 'ovs-docker', to integrate OVS with Docker
312 containers. If and when there is a native integration of Open vSwitch
313 with Docker, the wrapper script will be retired.
314 - Added support for DPDK Tunneling. VXLAN, GRE, and Geneve are supported
315 protocols. This is generic tunneling mechanism for userspace datapath.
316 - Support for multicast snooping (IGMPv1, IGMPv2 and IGMPv3)
317 - Support for Linux kernels up to 4.0.x
318 - The documentation now use the term 'destination' to mean one of syslog,
319 console or file for vlog logging instead of the previously used term
321 - Support for VXLAN Group Policy extension
322 - Initial support for the IETF Auto-Attach SPBM draft standard. This
323 contains rudimentary support for the LLDP protocol as needed for
325 - The default OpenFlow and OVSDB ports are now the IANA-assigned
326 numbers. OpenFlow is 6653 and OVSDB is 6640.
327 - Support for DPDK vHost.
328 - Support for outer UDP checksums in Geneve and VXLAN.
329 - The kernel vports with dependencies are no longer part of the overall
330 openvswitch.ko but built and loaded automatically as individual kernel
331 modules (vport-*.ko).
332 - Support for STT tunneling.
333 - ovs-sim: New developer tool for simulating multiple OVS instances.
334 See ovs-sim(1) for more information.
335 - Support to configure method (--syslog-method argument) that determines
336 how daemons will talk with syslog.
337 - Support for "ovs-appctl vlog/list-pattern" command that lets to query
338 logging message format for each destination.
342 ---------------------
343 - OpenFlow 1.1, 1.2, and 1.3 are now enabled by default in
345 - Linux kernel datapath now has an exact match cache optimizing the
346 flow matching process.
347 - Datapath flows now have partially wildcarded tranport port field
348 matches. This reduces userspace upcalls, but increases the
349 number of different masks in the datapath. The kernel datapath
350 exact match cache removes the overhead of matching the incoming
351 packets with the larger number of masks, but when paired with an
352 older kernel module, some workloads may perform worse with the
354 - Compatibility with autoconf 2.63 (previously >=2.64)
356 v2.2.0 - Internal Release
357 ---------------------
358 - Internal ports are no longer brought up by default, because it
359 should be an administrator task to bring up devices as they are
361 - ovs-vsctl now reports when ovs-vswitchd fails to create a new port or
363 - Port creation and configuration errors are now stored in a new error
364 column of the Interface table and included in 'ovs-vsctl show'.
365 - The "ovsdbmonitor" graphical tool has been removed, because it was
366 poorly maintained and not widely used.
367 - New "check-ryu" Makefile target for running Ryu tests for OpenFlow
368 controllers against Open vSwitch. See INSTALL.md for details.
369 - Added IPFIX support for SCTP flows and templates for ICMPv4/v6 flows.
370 - Upon the receipt of a SIGHUP signal, ovs-vswitchd no longer reopens its
371 log file (it will terminate instead). Please use 'ovs-appctl vlog/reopen'
373 - Support for Linux kernels up to 3.14. From Kernel 3.12 onwards OVS uses
374 tunnel API for GRE and VXLAN.
375 - Added DPDK support.
376 - Added support for custom vlog patterns in Python
380 ---------------------
381 - Address prefix tracking support for flow tables. New columns
382 "prefixes" in OVS-DB table "Flow_Table" controls which packet
383 header fields are used for address prefix tracking. Prefix
384 tracking allows the classifier to skip rules with longer than
385 necessary prefixes, resulting in better wildcarding for datapath
386 flows. Default configuration is to not use any fields for prefix
387 tracking. However, if any flow tables contain both exact matches
388 and masked matches for IP address fields, OVS performance may be
389 increased by using this feature.
390 * As of now, the fields for which prefix lookup can be enabled
391 are: 'tun_id', 'tun_src', 'tun_dst', 'nw_src', 'nw_dst' (or
392 aliases 'ip_src' and 'ip_dst'), 'ipv6_src', and 'ipv6_dst'.
393 (Using this feature for 'tun_id' would only make sense if the
394 tunnel IDs have prefix structure similar to IP addresses.)
395 * There is a maximum number of fields that can be enabled for any
396 one flow table. Currently this limit is 3.
398 $ ovs-vsctl set Bridge br0 flow_tables:0=@N1 -- \
399 --id=@N1 create Flow_Table name=table0
400 $ ovs-vsctl set Bridge br0 flow_tables:1=@N1 -- \
401 --id=@N1 create Flow_Table name=table1
402 $ ovs-vsctl set Flow_Table table0 prefixes=ip_dst,ip_src
403 $ ovs-vsctl set Flow_Table table1 prefixes=[]
404 - TCP flags matching: OVS now supports matching of TCP flags. This
405 has an adverse performance impact when using OVS userspace 1.10
406 or older (no megaflows support) together with the new OVS kernel
407 module. It is recommended that the kernel and userspace modules
408 both are upgraded at the same time.
409 - The default OpenFlow and OVSDB ports will change to
410 IANA-assigned numbers in a future release. Consider updating
411 your installations to specify port numbers instead of using the
414 * The OpenFlow 1.1+ "Write-Actions" instruction is now supported.
415 * OVS limits the OpenFlow port numbers it assigns to port 32767 and
416 below, leaving port numbers above that range free for assignment
418 * ovs-vswitchd now honors changes to the "ofport_request" column
419 in the Interface table by changing the port's OpenFlow port
421 * The Open vSwitch software switch now supports OpenFlow groups.
422 - ovs-vswitchd.conf.db.5 man page will contain graphviz/dot
423 diagram only if graphviz package was installed at the build time.
424 - Support for Linux kernels up to 3.11
426 The "show" command also displays mega flow mask stats.
428 * New command "ofp-parse-pcap" to dump OpenFlow from PCAP files.
429 - ovs-controller has been renamed test-controller. It is no longer
430 packaged or installed by default, because too many users assumed
431 incorrectly that ovs-controller was a necessary or desirable part
432 of an Open vSwitch deployment.
433 - Added vlog option to export to a UDP syslog sink.
435 * The "monitor" command can now monitor all tables in a database,
436 instead of being limited to a single table.
437 - The flow-eviction-threshold has been replaced by the flow-limit which is a
438 hard limit on the number of flows in the datapath. It defaults to 200,000
439 flows. OVS automatically adjusts this number depending on network
441 - Added IPv6 support for active and passive socket communications.
445 ---------------------
446 - The ovs-vswitchd process is no longer single-threaded. Multiple
447 threads are now used to handle flow set up and asynchronous
450 * Experimental support for OpenFlow 1.1 (in addition to 1.2 and
451 1.3, which had experimental support in 1.10).
452 * Experimental protocol support for OpenFlow 1.1+ groups. This
453 does not yet include an implementation in the Open vSwitch
455 * Experimental protocol support for OpenFlow 1.2+ meters. This
456 does not yet include an implementation in the Open vSwitch
458 * New support for matching outer source and destination IP address
459 of tunneled packets, for tunnel ports configured with the newly
460 added "remote_ip=flow" and "local_ip=flow" options.
461 * Support for matching on metadata 'pkt_mark' for interacting with
462 other system components. On Linux this corresponds to the skb
464 * Support matching, rewriting SCTP ports
465 - The Interface table in the database has a new "ifindex" column to
466 report the interface's OS-assigned ifindex.
467 - New "check-oftest" Makefile target for running OFTest against Open
468 vSwitch. See README-OFTest for details.
469 - The flow eviction threshold has been moved to the Open_vSwitch table.
470 - Database names are now mandatory when specifying ovsdb-server options
471 through database paths (e.g. Private key option with the database name
472 should look like "--private-key=db:Open_vSwitch,SSL,private_key").
473 - Added ovs-dev.py, a utility script helpful for Open vSwitch developers.
474 - Support for Linux kernels up to 3.10
476 * New "ofp-parse" for printing OpenFlow messages read from a file.
477 * New commands for OpenFlow 1.1+ groups.
478 - Added configurable flow caching support to IPFIX exporter.
479 - Dropped support for Linux pre-2.6.32.
480 - Log file timestamps and ovsdb commit timestamps are now reported
481 with millisecond resolution. (Previous versions only reported
485 v1.11.0 - 28 Aug 2013
486 ---------------------
487 - Support for megaflows, which allows wildcarding in the kernel (and
488 any dpif implementation that supports wildcards). Depending on
489 the flow table and switch configuration, flow set up rates are
490 close to the Linux bridge.
491 - The "tutorial" directory contains a new tutorial for some advanced
492 Open vSwitch features.
493 - Stable bond mode has been removed.
494 - The autopath action has been removed.
495 - New support for the data encapsulation format of the LISP tunnel
496 protocol (RFC 6830). An external control plane or manual flow
497 setup is required for EID-to-RLOC mapping.
499 * The "dec_mpls_ttl" and "set_mpls_ttl" actions from OpenFlow
500 1.1 and later are now implemented.
501 * New "stack" extension for use in actions, to push and pop from
503 * The "load" and "set_field" actions can now modify the "in_port". (This
504 allows one to enable output to a flow's input port by setting the
505 in_port to some unused value, such as OFPP_NONE.)
507 * New debugging commands "add-flow", "mod-flow", "del-flow".
508 * "dump-flows" now has a -m option to increase output verbosity.
509 - In dpif-based bridges, cache action translations, which can improve
510 flow set up performance by 80% with a complicated flow table.
511 - New syslog format, prefixed with "ovs|", to be easier to filter.
512 - RHEL: Removes the default firewall rule that allowed GRE traffic to
513 pass through. Any users that relied on this automatic firewall hole
514 will have to manually configure it. The ovs-ctl(8) manpage documents
515 the "enable-protocol" command that can be used as an alternative.
516 - New CFM demand mode which uses data traffic to indicate interface
519 v1.10.0 - 01 May 2013
520 ---------------------
521 - Bridge compatibility support has been removed. Any uses that
522 rely on ovs-brcompatd will have to stick with Open vSwitch 1.9.x
523 or adapt to native Open vSwitch support (e.g. use ovs-vsctl instead
525 - The maximum size of the MAC learning table is now configurable.
526 - With the Linux datapath, packets for new flows are now queued
527 separately on a per-port basis, so it should no longer be
528 possible for a large number of new flows arriving on one port to
529 prevent new flows from being processed on other ports.
531 * Previously ovs-vsctl would retry connecting to the database forever,
532 causing it to hang if ovsdb-server was not running. Now, ovs-vsctl
533 only tries once by default (use --retry to try forever). This change
534 means that you may want to remove uses of --timeout to avoid hangs
536 * Many "ovs-vsctl" database commands now accept an --if-exists option.
537 Please refer to the ovs-vsctl manpage for details.
539 - Experimental support for newer versions of OpenFlow. See
540 the "What versions of OpenFlow does Open vSwitch support?"
541 question in the FAQ for more details.
542 - The OpenFlow "dp_desc" may now be configured by setting the
543 value of other-config:dp-desc in the Bridge table.
544 - It is possible to request the OpenFlow port number with the
545 "ofport_request" column in the Interface table.
546 - The NXM flow_removed message now reports the OpenFlow table ID
547 from which the flow was removed.
549 - New support for the VXLAN tunnel protocol (see the IETF draft here:
550 http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03).
551 - Tunneling requires the version of the kernel module paired with
552 Open vSwitch 1.9.0 or later.
553 - Inheritance of the Don't Fragment bit in IP tunnels (df_inherit)
554 is no longer supported.
555 - Path MTU discovery is no longer supported.
556 - CAPWAP tunneling support removed.
557 - Tunnels with multicast destination ports are no longer supported.
559 - The "dump-flows" and "del-flows" no longer require an argument
560 if only one datapath exists.
562 - New "vlog/disable-rate-limit" and "vlog/enable-rate-limit"
563 commands available allow control over logging rate limits.
564 - New "dpif/dump-dps", "dpif/show", and "dpif/dump-flows" command
565 that mimic the equivalent ovs-dpctl commands.
566 - The ofproto library is now responsible for assigning OpenFlow port
567 numbers. An ofproto implementation should assign them when
568 port_construct() is called.
569 - All dpif-based bridges of a particular type share a common
570 datapath called "ovs-<type>", e.g. "ovs-system". The ovs-dpctl
571 commands will now return information on that shared datapath. To
572 get the equivalent bridge-specific information, use the new
573 "ovs-appctl dpif/*" commands.
574 - Backward-incompatible changes:
575 - Earlier Open vSwitch versions treated ANY as a wildcard in flow
576 syntax. OpenFlow 1.1 adds a port named ANY, which introduces a
577 conflict. ANY was rarely used in flow syntax, so we chose to
578 retire that meaning of ANY in favor of the OpenFlow 1.1 meaning.
579 - Patch ports no longer require kernel support, so they now work
580 with FreeBSD and the kernel module built into Linux 3.3 and later.
581 - New "sample" action.
585 ------------------------
587 - Support for ipv6 set action.
588 - SKB mark matching and setting.
589 - support for Linux kernels up to 3.8
590 - FreeBSD is now a supported platform, thanks to code contributions from
591 Gaetano Catalli, Ed Maste, and Giuseppe Lettieri.
592 - ovs-bugtool: New --ovs option to report only OVS related information.
593 - New %t and %T log escapes to identify the subprogram within a
594 cooperating group of processes or threads that emitted a log message.
595 The default log patterns now include this information.
597 - Allow bitwise masking for SHA and THA fields in ARP, SLL and TLL
598 fields in IPv6 neighbor discovery messages, and IPv6 flow label.
599 - Adds support for writing to the metadata field for a flow.
601 - The tunneling code no longer assumes input and output keys are
602 symmetric. If they are not, PMTUD needs to be disabled for
603 tunneling to work. Note this only applies to flow-based keys.
604 - New support for a nonstandard form of GRE that supports a 64-bit key.
605 - Tunnel Path MTU Discovery default value was set to 'disabled'.
606 This feature is deprecated and will be removed soon.
607 - Tunnel header caching removed.
609 - Commands and actions that accept port numbers now also accept keywords
610 that represent those ports (such as LOCAL, NONE, and ALL). This is
611 also the recommended way to specify these ports, for compatibility
612 with OpenFlow 1.1 and later (which use the OpenFlow 1.0 numbers
613 for these ports for different purposes).
615 - Support requesting the port number with the "port_no" option in
616 the "add-if" command.
617 - ovs-pki: The "online PKI" features have been removed, along with
618 the ovs-pki-cgi program that facilitated it, because of some
619 alarmist insecurity claims. We do not believe that these claims
620 are true, but because we do not know of any users for this
621 feature it seems better on balance to remove it. (The ovs-pki-cgi
622 program was not included in distribution packaging.)
623 - ovsdb-server now enforces the immutability of immutable columns. This
624 was not enforced in earlier versions due to an oversight.
625 - The following features are now deprecated. They will be removed no
626 earlier than February 2013. Please email dev@openvswitch.org with
628 - Bridge compatibility.
630 - The autopath action.
631 - Interface type "null".
632 - Numeric values for reserved ports (see "ovs-ofctl" note above).
633 - Tunnel Path MTU Discovery.
634 - CAPWAP tunnel support.
635 - The data in the RARP packets can now be matched in the same way as the
640 ------------------------
641 *** Internal only release ***
642 - New FAQ. Please send updates and additions!
643 - Authors of controllers, please read the new section titled "Action
644 Reproduction" in DESIGN, which describes an Open vSwitch change in
645 behavior in corner cases that may affect some controllers.
647 - A new test utility that can create L3 tunnel between two Open
648 vSwitches and detect connectivity issues.
650 - New --sort and --rsort options for "dump-flows" command.
651 - "mod-port" command can now control all OpenFlow config flags.
653 - Allow general bitwise masking for IPv4 and IPv6 addresses in
654 IPv4, IPv6, and ARP packets. (Previously, only CIDR masks
656 - Allow support for arbitrary Ethernet masks. (Previously, only
657 the multicast bit in the destination address could be individually
659 - New field OXM_OF_METADATA, to align with OpenFlow 1.1.
660 - The OFPST_QUEUE request now reports an error if a specified port or
661 queue does not exist, or for requests for a specific queue on all
662 ports, if the specified queue does not exist on any port. (Previous
663 versions generally reported an empty set of results.)
664 - New "flow monitor" feature to allow controllers to be notified of
665 flow table changes as they happen.
666 - Additional protocols are not mirrored and dropped when forward-bpdu is
667 false. For a full list, see the ovs-vswitchd.conf.db man page.
668 - Open vSwitch now sends RARP packets in situations where it previously
669 sent a custom protocol, making it consistent with behavior of QEMU and
671 - All Open vSwitch programs and log files now show timestamps in UTC,
672 instead the local timezone, by default.
676 ------------------------
677 - kernel modules are renamed. openvswitch_mod.ko is now
678 openvswitch.ko and brcompat_mod.ko is now brcompat.ko.
679 - Increased the number of NXM registers to 8.
680 - Added ability to configure DSCP setting for manager and controller
681 connections. By default, these connections have a DSCP value of
682 Internetwork Control (0xc0).
683 - Added the granular link health statistics, 'cfm_health', to an
686 - Added support to mask nd_target for ICMPv6 neighbor discovery flows.
687 - Added support for OpenFlow 1.3 port description (OFPMP_PORT_DESC)
690 - Added the "dump-ports-desc" command to retrieve port
691 information using the new port description multipart messages.
693 - Added support for spawning ovs-test server from the client.
694 - Now ovs-test is able to automatically create test bridges and ports.
695 - "ovs-dpctl dump-flows" now prints observed TCP flags in TCP flows.
696 - Tripled flow setup performance.
697 - The "coverage/log" command previously available through ovs-appctl
698 has been replaced by "coverage/show". The new command replies with
699 coverage counter values, instead of logging them.
703 ------------------------
704 - Allow OFPP_CONTROLLER as the in_port for packet-out messages.
708 ------------------------
709 *** Internal only release ***
711 - LACP bonds no longer fall back to balance-slb when negotiations fail.
712 Instead they drop traffic.
713 - The default bond_mode changed from SLB to active-backup, to protect
714 unsuspecting users from the significant risks of SLB bonds (which are
715 documented in vswitchd/INTERNALS).
716 - Load balancing can be disabled by setting the bond-rebalance-interval
719 - Added support for bitwise matching on TCP and UDP ports.
720 See ovs-ofctl(8) for more information.
721 - NXM flow dumps now include times elapsed toward idle and hard
723 - Added an OpenFlow extension NXT_SET_ASYNC_CONFIG that allows
724 controllers more precise control over which OpenFlow messages they
725 receive asynchronously.
726 - New "fin_timeout" action.
727 - Added "fin_timeout" support to "learn" action.
728 - New Nicira action NXAST_CONTROLLER that offers additional features
729 over output to OFPP_CONTROLLER.
730 - When QoS settings for an interface do not configure a default queue
731 (queue 0), Open vSwitch now uses a default configuration for that
732 queue, instead of dropping all packets as in previous versions.
734 - Logging to console and file will have UTC timestamp as a default for
735 all the daemons. An example of the default format is
736 2012-01-27T16:35:17Z. ovs-appctl can be used to change the default
738 - The syntax of commands and options to set log levels was simplified,
739 to make it easier to remember.
740 - New support for limiting the number of flows in an OpenFlow flow
741 table, with configurable policy for evicting flows upon
742 overflow. See the Flow_Table table in ovs-vswitch.conf.db(5)
743 for more information.
744 - New "enable-async-messages" column in the Controller table. If set to
745 false, OpenFlow connections to the controller will initially have all
746 asynchronous messages disabled, overriding normal OpenFlow behavior.
747 - ofproto-provider interface:
748 - "struct rule" has a new member "used" that ofproto implementations
749 should maintain by updating with ofproto_rule_update_used().
751 - The new option --timestamp causes the "monitor" command to print
752 a timestamp with every update.
753 - CFM module CCM broadcasts can now be tagged with an 802.1p priority.
757 ------------------------
759 - Added support for querying, modifying, and deleting flows
760 based on flow cookie when using NXM.
761 - Added new NXM_PACKET_IN format.
762 - Added new NXAST_DEC_TTL action.
764 - Added daemonization support to the monitor and snoop commands.
766 - The "find" command supports new set relational operators
767 {=}, {!=}, {<}, {>}, {<=}, and {>=}.
768 - ovsdb-tool now uses the typical database and schema installation
769 directories as defaults.
770 - The default MAC learning timeout has been increased from 60 seconds
771 to 300 seconds. The MAC learning timeout is now configurable.
775 ------------------------
776 - Compatible with Open vSwitch kernel module included in Linux 3.3.
777 - New "VLAN splinters" feature to work around buggy device drivers
778 in old Linux versions. (This feature is deprecated. When
779 broken device drivers are no longer in widespread use, we will
780 delete this feature.) See ovs-vswitchd.conf.db(5) for more
783 - Added ability to match on IPv6 flow label through NXM.
784 - Added ability to match on ECN bits in IPv4 and IPv6 through NXM.
785 - Added ability to match on TTL in IPv4 and IPv6 through NXM.
786 - Added ability to modify ECN bits in IPv4.
787 - Added ability to modify TTL in IPv4.
789 - Don't require the "normal" action to use mirrors. Traffic will
790 now be properly mirrored for any flows, regardless of their
792 - Track packet and byte statistics sent on mirrors.
793 - The sFlow implementation can now usually infer the correct agent
794 device instead of having to be told explicitly.
796 - New "fdb/flush" command to flush bridge's MAC learning table.
798 - A new distributed testing tool that allows one to diagnose performance
799 and connectivity issues. This tool currently is not included in RH or
801 - RHEL packaging now supports integration with Red Hat network scripts.
803 - Post 1.4.*, OVS will be changing the default bond mode from balance-slb
804 to active-backup. SLB bonds carry significant risks with them
805 (documented vswitchd/INTERNALS) which we want to prevent unsuspecting
806 users from running into. Users are advised to update any scripts or
807 configuration which may be negatively impacted by explicitly setting
808 the bond mode which they want to use.
812 ------------------------
814 - Added an OpenFlow extension which allows the "output" action to accept
816 - Added an OpenFlow extension for flexible learning.
817 - Bumped number of NXM registers from four to five.
819 - New "version" command to determine version of running daemon.
820 - If no argument is provided for "cfm/show", displays detailed
821 information about all interfaces with CFM enabled.
822 - If no argument is provided for "lacp/show", displays detailed
823 information about all ports with LACP enabled.
825 - New "set-if" command to modify a datapath port's configuration.
827 - The software switch now supports 255 OpenFlow tables, instead
828 of just one. By default, only table 0 is consulted, but the
829 new NXAST_RESUBMIT_TABLE action can look up in additional
830 tables. Tables 128 and above are reserved for use by the
831 switch itself; please use only tables 0 through 127.
832 - Add support for 802.1D spanning tree (STP).
833 - Fragment handling extensions:
834 - New OFPC_FRAG_NX_MATCH fragment handling mode, in which L4
835 fields are made available for matching in fragments with
837 - New NXM_NX_IP_FRAG match field for matching IP fragments (usable
838 via "ip_frag" in ovs-ofctl).
839 - New ovs-ofctl "get-frags" and "set-frags" commands to get and set
840 fragment handling policy.
841 - CAPWAP tunneling now supports an extension to transport a 64-bit key.
842 By default it remains compatible with the old version and other
843 standards-based implementations.
844 - Flow setups are now processed in a round-robin manner across ports
845 to prevent any single client from monopolizing the CPU and conducting
846 a denial of service attack.
847 - Added support for native VLAN tagging. A new "vlan_mode"
848 parameter can be set for "port". Possible values: "access",
849 "trunk", "native-tagged" and "native-untagged".
850 - test-openflowd has been removed. Please use ovs-vswitchd instead.
853 ------------------------
854 - New "ofproto" abstraction layer to ease porting to hardware
856 - Packaging for Red Hat Enterprise Linux 5.6 and 6.0.
857 - Datapath support for Linux kernels up to 3.0.
859 - New "bundle" and "bundle_load" action extensions.
861 - Implement table unique constraints.
862 - Support cooperative locking between callers.
864 - New "-s" option for "show" command prints packet and byte
865 counters for each port.
867 - New "--readd" option for "replace-flows".
869 - New "show" command to print an overview of configuration.
870 - New "comment" command to add remark that explains intentions.
871 - ovs-brcompatd has been rewritten to fix long-standing bugs.
872 - ovs-openflowd has been renamed test-openflowd and moved into the
873 tests directory. Its presence confused too many users. Please
874 use ovs-vswitchd instead.
875 - New ovs-benchmark utility to test flow setup performance.
876 - A new log level "off" has been added. Configuring a log facility
877 "off" prevents any messages from being logged to it. Previously,
878 "emer" was effectively "off" because no messages were ever logged at
879 level "emer". Now, errors that cause a process to exit are logged
881 - "configure" option --with-l26 has been renamed --with-linux, and
882 --with-l26-source has been renamed --with-linux-source. The old
883 names will be removed after the next release, so please update
885 - The "-2.6" suffix has been dropped from the datapath/linux-2.6 and
886 datapath/linux-2.6/compat-2.6 directories.
888 - Dropped support for "tun_id_from_cookie" OpenFlow extension.
889 Please use the extensible match extensions instead.
890 - Removed the Maintenance_Point and Monitor tables in an effort
891 to simplify 802.1ag configuration.
892 - Performance and scalability improvements
896 ------------------------
897 - Ability to define policies over IPv6
900 - Support for extensible match extensions to OpenFlow
902 - Support for HFSC qdisc.
903 - Queue used by in-band control can now be configured.
905 - Kernel<->userspace interface has been reworked and should be
906 close to a stable ABI now.
907 - "Port group" concept has been dropped.
908 - GRE over IPSEC tunnels
910 - New active backup bonding mode.
911 - New L4 hashing support when LACP is enabled.
912 - Source MAC hash now includes VLAN field also.
914 - Greatly improved handling of large flow tables
916 - "show" command now prints full vport configuration.
917 - "dump-groups" command removed since kernel support for
918 port groups was dropped.
920 - New commands for working with the new Managers table.
921 - "list" command enhanced with new formatting options and --columns
923 - "get" command now accepts new --id option.
924 - New "find" command.
926 - New "queue-stats" command for printing queue stats.
927 - New commands "replace-flows" and "diff-flows".
928 - Commands to add and remove flows can now read from files.
929 - New --flow-format option to enable or disable NXM.
930 - New --more option to increase OpenFlow message verbosity.
931 - Removed "tun-cookie" command, which is no longer useful.
932 - ovs-controller enhancements for testing various features.
933 - New ovs-vlan-test command for testing for Linux kernel driver VLAN
934 bugs. New ovs-vlan-bug-workaround command for enabling and
935 disabling a workaround for these driver bugs.
937 - "Resubmit" actions now update flow statistics.
938 - New "register" extension for use in matching and actions, via NXM.
939 - New "multipath" experimental action extension.
940 - New support for matching multicast Ethernet frames, via NXM.
941 - New extension for OpenFlow vendor error codes.
942 - New extension to set the QoS output queue without actually
943 sending to an output port.
944 - Open vSwitch now reports a single flow table, instead of
945 separate hash and wildcard tables. This better models the
946 current implementation.
947 - New experimental "note" action.
948 - New "ofproto/trace" ovs-appctl command and associated utilities
949 to ease debugging complex flow tables.
951 - Schema documentation now includes an entity-relationship diagram.
952 - The database is now garbage collected. In most tables,
953 unreferenced rows will be deleted automatically.
954 - Many tables now include statistics updated periodically by
955 ovs-vswitchd or ovsdb-server.
956 - Every table now has an "external-ids" column for use by OVS
958 - There is no default controller anymore. Each bridge must have its
959 controller individually specified.
960 - The "fail-mode" is now a property of a Bridge instead of a Controller.
961 - New versioning and checksum features.
962 - New Managers table and manager_options column in Open_vSwitch table
963 for specifying managers. The old "managers" column in the
964 Open_vSwitch table has been removed.
965 - Many "name" columns are now immutable.
967 - Dropped support for XenServer pre-5.6.100.
968 - Dropped support for Linux pre-2.6.18.
969 - Dropped controller discovery support.
970 - Dropped "ovs-ofctl status" and the OpenFlow extension that it used.
971 Statistics reporting in the database is a rough equivalent.
972 - Dropped the "corekeeper" package (now separate, at
973 http://openvswitch.org/cgi-bin/gitweb.cgi?p=corekeeper).
974 - Performance and scalability improvements
977 v1.1.0pre2 - 13 Sep 2010
978 ------------------------
981 v1.1.0pre1 - 31 Aug 2010
982 ------------------------
983 - OpenFlow 1.0 slicing (QoS) functionality
984 - Python bindings for configuration database (no write support)
985 - Performance and scalability improvements
990 - New "patch" interface type
995 - Configuration database with remote management
998 - Support for XenServer 5.5 and 5.6
999 - Performance and scalability improvements
1002 v0.99.2 - 18 Feb 2010
1003 ---------------------
1006 v0.99.1 - 25 Jan 2010
1007 ---------------------
1008 - Add support for sFlow(R)
1009 - Make headers compatible with C++
1012 v0.99.0 - 14 Jan 2010
1013 ---------------------
1014 - User-space forwarding engine
1017 v0.90.7 - 29 Nov 2009
1018 ---------------------
1019 - Add support for NetFlow active timeouts
1022 v0.90.6 - 6 Oct 2009
1023 --------------------
1026 v0.90.5 - 21 Sep 2009
1027 ---------------------
1028 - Generalize in-band control to more diverse network setups
projects / mirror_ovs.git / blob