ovs-ofctl(8) Open vSwitch Manual ovs-ofctl(8)
NAME
ovs-ofctl - administer OpenFlow switches
SYNOPSIS
ovs-ofctl [options] command [switch] [args...]
DESCRIPTION
The ovs-ofctl program is a command line tool for monitoring and admin‐
istering OpenFlow switches. It can also show the current state of an
OpenFlow switch, including features, configuration, and table entries.
It should work with any OpenFlow switch, not just Open vSwitch.
OpenFlow Switch Management Commands
These commands allow ovs-ofctl to monitor and administer an OpenFlow
switch. It is able to show the current state of a switch, including
features, configuration, and table entries.
Most of these commands take an argument that specifies the method for
connecting to an OpenFlow switch. The following connection methods are
supported:
ssl:ip[:port]
tcp:ip[:port]
The specified port on the host at the given ip, which
must be expressed as an IP address (not a DNS name) in
IPv4 or IPv6 address format. Wrap IPv6 addresses in
square brackets, e.g. tcp:[::1]:6653. For ssl, the
--private-key, --certificate, and --ca-cert options are
mandatory.
If port is not specified, it defaults to 6653.
unix:file
On POSIX, a Unix domain server socket named file.
On Windows, a localhost TCP port written in file.
file This is short for unix:file, as long as file does not
contain a colon.
bridge This is short for unix:/var/run/openvswitch/bridge.mgmt,
as long as bridge does not contain a colon.
[type@]dp
Attempts to look up the bridge associated with dp and
open as above. If type is given, it specifies the data‐
path provider of dp, otherwise the default provider sys‐
tem is assumed.
show switch
Prints to the console information on switch, including informa‐
tion on its flow tables and ports.
dump-tables switch
Prints to the console statistics for each of the flow tables
used by switch.
dump-table-features switch
Prints to the console features for each of the flow tables used
by switch.
dump-table-desc switch
Prints to the console configuration for each of the flow tables
used by switch for OpenFlow 1.4+.
mod-table switch table_id setting
This command configures flow table settings for OpenFlow table
table_id within switch. The available settings depend on the
OpenFlow version in use. In OpenFlow 1.1 and 1.2 (which must be
enabled with the -O option) only, mod-table configures behavior
when no flow is found when a packet is looked up in a flow ta‐
ble. The following setting values are available:
drop Drop the packet.
continue
Continue to the next table in the pipeline. (This is how
an OpenFlow 1.0 switch always handles packets that do not
match any flow, in tables other than the last one.)
controller
Send to controller. (This is how an OpenFlow 1.0 switch
always handles packets that do not match any flow in the
last table.)
In OpenFlow 1.4 and later (which must be enabled with the -O
option) only, mod-table configures the behavior when a con‐
troller attempts to add a flow to a flow table that is full.
The following setting values are available:
evict Delete some existing flow from the flow table, according
to the algorithm described for the Flow_Table table in
ovs-vswitchd.conf.db(5).
noevict
Refuse to add the new flow. (Eviction might still be
enabled through the overflow_policy oclumn in the
Flow_Table table documented in ovs-vswitchd.conf.db(5).)
dump-ports switch [netdev]
Prints to the console statistics for network devices associated
with switch. If netdev is specified, only the statistics asso‐
ciated with that device will be printed. netdev can be an Open‐
Flow assigned port number or device name, e.g. eth0.
dump-ports-desc switch [port]
Prints to the console detailed information about network devices
associated with switch. To dump only a specific port, specify
its number as port. Otherwise, if port is omitted, or if it is
specified as ANY, then all ports are printed. This is a subset
of the information provided by the show command.
If the connection to switch negotiates OpenFlow 1.0, 1.2, or
1.2, this command uses an OpenFlow extension only implemented in
Open vSwitch (version 1.7 and later).
Only OpenFlow 1.5 and later support dumping a specific port.
Earlier versions of OpenFlow always dump all ports.
mod-port switch port action
Modify characteristics of port port in switch. port may be an
OpenFlow port number or name or the keyword LOCAL (the preferred
way to refer to the OpenFlow local port). The action may be any
one of the following:
up
down Enable or disable the interface. This is equivalent to
ifconfig up or ifconfig down on a Unix system.
stp
no-stp Enable or disable 802.1D spanning tree protocol (STP) on
the interface. OpenFlow implementations that don't sup‐
port STP will refuse to enable it.
receive
no-receive
receive-stp
no-receive-stp
Enable or disable OpenFlow processing of packets received
on this interface. When packet processing is disabled,
packets will be dropped instead of being processed
through the OpenFlow table. The receive or no-receive
setting applies to all packets except 802.1D spanning
tree packets, which are separately controlled by
receive-stp or no-receive-stp.
forward
no-forward
Allow or disallow forwarding of traffic to this inter‐
face. By default, forwarding is enabled.
flood
no-flood
Controls whether an OpenFlow flood action will send traf‐
fic out this interface. By default, flooding is enabled.
Disabling flooding is primarily useful to prevent loops
when a spanning tree protocol is not in use.
packet-in
no-packet-in
Controls whether packets received on this interface that
do not match a flow table entry generate a ``packet in''
message to the OpenFlow controller. By default, ``packet
in'' messages are enabled.
The show command displays (among other information) the configu‐
ration that mod-port changes.
get-frags switch
Prints switch's fragment handling mode. See set-frags, below,
for a description of each fragment handling mode.
The show command also prints the fragment handling mode among
its other output.
set-frags switch frag_mode
Configures switch's treatment of IPv4 and IPv6 fragments. The
choices for frag_mode are:
normal Fragments pass through the flow table like non-fragmented
packets. The TCP ports, UDP ports, and ICMP type and
code fields are always set to 0, even for fragments where
that information would otherwise be available (fragments
with offset 0). This is the default fragment handling
mode for an OpenFlow switch.
drop Fragments are dropped without passing through the flow
table.
reassemble
The switch reassembles fragments into full IP packets
before passing them through the flow table. Open vSwitch
does not implement this fragment handling mode.
nx-match
Fragments pass through the flow table like non-fragmented
packets. The TCP ports, UDP ports, and ICMP type and
code fields are available for matching for fragments with
offset 0, and set to 0 in fragments with nonzero offset.
This mode is a Nicira extension.
See the description of ip_frag, below, for a way to match on
whether a packet is a fragment and on its fragment offset.
dump-flows switch [flows]
Prints to the console all flow entries in switch's tables that
match flows. If flows is omitted, all flows in the switch are
retrieved. See Flow Syntax, below, for the syntax of flows.
The output format is described in Table Entry Output.
By default, ovs-ofctl prints flow entries in the same order that
the switch sends them, which is unlikely to be intuitive or con‐
sistent. See the description of --sort and --rsort, under
OPTIONS below, to influence the display order.
dump-aggregate switch [flows]
Prints to the console aggregate statistics for flows in switch's
tables that match flows. If flows is omitted, the statistics
are aggregated across all flows in the switch's flow tables.
See Flow Syntax, below, for the syntax of flows. The output
format is described in Table Entry Output.
queue-stats switch [port [queue]]
Prints to the console statistics for the specified queue on port
within switch. port can be an OpenFlow port number or name, the
keyword LOCAL (the preferred way to refer to the OpenFlow local
port), or the keyword ALL. Either of port or queue or both may
be omitted (or equivalently the keyword ALL). If both are omit‐
ted, statistics are printed for all queues on all ports. If
only queue is omitted, then statistics are printed for all
queues on port; if only port is omitted, then statistics are
printed for queue on every port where it exists.
OpenFlow 1.1+ Group Table Commands
The following commands work only with switches that support OpenFlow
1.1 or later. Because support for OpenFlow 1.1 and later is still
experimental in Open vSwitch, it is necessary to explicitly enable
these protocol versions in ovs-ofctl (using -O) and in the switch
itself (with the protocols column in the Bridge table). For more
information, see ``Q: What versions of OpenFlow does Open vSwitch sup‐
port?'' in the Open vSwitch FAQ.
dump-groups switch [group]
Prints group entries in switch's tables to console. To dump
only a specific group, specify its number as group. Otherwise,
if group is omitted, or if it is specified as ALL, then all
groups are printed. Each line of output is a group entry as
described in Group Syntax below.
Only OpenFlow 1.5 and later support dumping a specific group.
Earlier versions of OpenFlow always dump all groups.
dump-group-features switch
Prints to the console the group features of the switch.
dump-group-stats switch [groups]
Prints to the console statistics for the specified groups in the
switch's tables. If groups is omitted then statistics for all
groups are printed. See Group Syntax, below, for the syntax of
groups.
OpenFlow 1.3+ Switch Meter Table Commands
These commands manage the meter table in an OpenFlow switch. In each
case, meter specifies a meter entry in the format described in Meter
Syntax, below.
OpenFlow 1.3 introduced support for meters, so these commands only work
with switches that support OpenFlow 1.3 or later. The caveats
described for groups in the previous section also apply to meters.
add-meter switch meter
Add a meter entry to switch's tables. The meter syntax is
described in section Meter Syntax, below.
mod-meter switch meter
Modify an existing meter.
del-meters switch
del-meter switch [meter]
Delete entries from switch's meter table. meter can specify a
single meter with syntax meter=id, or all meters with syntax
meter=all.
dump-meters switch
dump-meter switch [meter]
Print meter configuration. meter can specify a single meter
with syntax meter=id, or all meters with syntax meter=all.
meter-stats switch [meter]
Print meter statistics. meter can specify a single meter with
syntax meter=id, or all meters with syntax meter=all.
meter-features switch
Print meter features.
OpenFlow Switch Flow Table Commands
These commands manage the flow table in an OpenFlow switch. In each
case, flow specifies a flow entry in the format described in Flow Syn‐
tax, below, file is a text file that contains zero or more flows in the
same syntax, one per line, and the optional --bundle option operates
the command as a single atomic transation, see option --bundle, below.
[--bundle] add-flow switch flow
[--bundle] add-flow switch - file
[--bundle] add-flows switch file
Add each flow entry to switch's tables. Each flow specification
(e.g., each line in file) may start with add, modify, delete,
modify_strict, or delete_strict keyword to specify whether a
flow is to be added, modified, or deleted, and whether the mod‐
ify or delete is strict or not. For backwards compatibility a
flow specification without one of these keywords is treated as a
flow add. All flow mods are executed in the order specified.
[--bundle] [--strict] mod-flows switch flow
[--bundle] [--strict] mod-flows switch - file
Modify the actions in entries from switch's tables that match
the specified flows. With --strict, wildcards are not treated
as active for matching purposes.
[--bundle] del-flows switch
[--bundle] [--strict] del-flows switch [flow]
[--bundle] [--strict] del-flows switch - file
Deletes entries from switch's flow table. With only a switch
argument, deletes all flows. Otherwise, deletes flow entries
that match the specified flows. With --strict, wildcards are
not treated as active for matching purposes.
[--bundle] [--readd] replace-flows switch file
Reads flow entries from file (or stdin if file is -) and queries
the flow table from switch. Then it fixes up any differences,
adding flows from flow that are missing on switch, deleting
flows from switch that are not in file, and updating flows in
switch whose actions, cookie, or timeouts differ in file.
With --readd, ovs-ofctl adds all the flows from file, even those
that exist with the same actions, cookie, and timeout in switch.
This resets all the flow packet and byte counters to 0, which
can be useful for debugging.
diff-flows source1 source2
Reads flow entries from source1 and source2 and prints the dif‐
ferences. A flow that is in source1 but not in source2 is
printed preceded by a -, and a flow that is in source2 but not
in source1 is printed preceded by a +. If a flow exists in both
source1 and source2 with different actions, cookie, or timeouts,
then both versions are printed preceded by - and +, respec‐
tively.
source1 and source2 may each name a file or a switch. If a name
begins with / or ., then it is considered to be a file name. A
name that contains : is considered to be a switch. Otherwise,
it is a file if a file by that name exists, a switch if not.
For this command, an exit status of 0 means that no differences
were found, 1 means that an error occurred, and 2 means that
some differences were found.
packet-out switch in_port actions packet...
Connects to switch and instructs it to execute the OpenFlow
actions on each packet. Each packet is specified as a series of
hex digits. For the purpose of executing the actions, the pack‐
ets are considered to have arrived on in_port, which may be an
OpenFlow port number or name (e.g. eth0), the keyword LOCAL (the
preferred way to refer to the OpenFlow ``local'' port), or the
keyword NONE to indicate that the packet was generated by the
switch itself.
OpenFlow Switch Group Table Commands
These commands manage the group table in an OpenFlow switch. In each
case, group specifies a group entry in the format described in Group
Syntax, below, and file is a text file that contains zero or more
groups in the same syntax, one per line.
add-group switch group
add-group switch - file
add-groups switch file
Add each group entry to switch's tables.
mod-group switch group
mod-group switch - file
Modify the action buckets in entries from switch's tables for
each group entry.
del-groups switch
del-groups switch [group]
del-groups switch - file
Deletes entries from switch's group table. With only a switch
argument, deletes all groups. Otherwise, deletes the group for
each group entry.
insert-buckets switch group
insert-buckets switch - file
Add buckets to an existing group present in the switch's group
table. If no command_bucket_id is present in the group specifi‐
cation then all buckets of the group are removed.
remove-buckets switch group
remove-buckets switch - file
Remove buckets to an existing group present in the switch's
group table. If no command_bucket_id is present in the group
specification then all buckets of the group are removed.
OpenFlow Switch Geneve Option Table Commands
Open vSwitch maintains a mapping table between Geneve options (defined
by gt;) and an NXM field tun_metadatan, where n
ranges from 0 to 63, that can be operated on for the purposes of
matches, actions, etc. This mapping must be explicitly specified by the
user through the following commands.
A Geneve option mapping is specified with the syntax
{class=class,type=type,len=length}->gt;>gt;tun_metadatan. When an option map‐
ping exists for a given tun_metadatan, matching on the defined field
becomes possible, e.g.:
ovs-ofctl add-geneve-map br0
"{class=0xffff,type=0,len=4}->gt;tun_metadata0"
ovs-ofctl add-flow br0 tun_metadata0=1234,actions=controller
A mapping should not be changed while it is in active use by a flow.
The result of doing so is undefined.
Currently, the Geneve mapping table is shared between all OpenFlow
switches in a given instance of Open vSwitch. This restriction will be
lifted in the future to allow for easier management.
These commands are Nicira extensions to OpenFlow and require Open
vSwitch 2.5 or later.
add-geneve-map switch option[,option]...
Add each option to switch's tables. Duplicate fields are
rejected.
del-geneve-map switch [option[,option]]...
Delete each option from switch's table, or all Geneve option
mapping if no option is specified. Fields that aren't mapped
are ignored.
dump-geneve-map switch
Show the currently mapped fields in the switch's option table as
well as switch capabilities.
OpenFlow Switch Monitoring Commands
snoop switch
Connects to switch and prints to the console all OpenFlow mes‐
sages received. Unlike other ovs-ofctl commands, if switch is
the name of a bridge, then the snoop command connects to a Unix
domain socket named /var/run/openvswitch/switch.snoop.
ovs-vswitchd listens on such a socket for each bridge and sends
to it all of the OpenFlow messages sent to or received from its
configured OpenFlow controller. Thus, this command can be used
to view OpenFlow protocol activity between a switch and its con‐
troller.
When a switch has more than one controller configured, only the
traffic to and from a single controller is output. If none of
the controllers is configured as a master or a slave (using a
Nicira extension to OpenFlow 1.0 or 1.1, or a standard request
in OpenFlow 1.2 or later), then a controller is chosen arbitrar‐
ily among them. If there is a master controller, it is chosen;
otherwise, if there are any controllers that are not masters or
slaves, one is chosen arbitrarily; otherwise, a slave controller
is chosen arbitrarily. This choice is made once at connection
time and does not change as controllers reconfigure their roles.
If a switch has no controller configured, or if the configured
controller is disconnected, no traffic is sent, so monitoring
will not show any traffic.
monitor switch [miss-len] [invalid_ttl] [watch:[spec...]]
Connects to switch and prints to the console all OpenFlow mes‐
sages received. Usually, switch should specify the name of a
bridge in the ovs-vswitchd database.
If miss-len is provided, ovs-ofctl sends an OpenFlow ``set con‐
figuration'' message at connection setup time that requests
miss-len bytes of each packet that misses the flow table. Open
vSwitch does not send these and other asynchronous messages to
an ovs-ofctl monitor client connection unless a nonzero value is
specified on this argument. (Thus, if miss-len is not speci‐
fied, very little traffic will ordinarily be printed.)
If invalid_ttl is passed, ovs-ofctl sends an OpenFlow ``set con‐
figuration'' message at connection setup time that requests
INVALID_TTL_TO_CONTROLLER, so that ovs-ofctl monitor can receive
``packet-in'' messages when TTL reaches zero on dec_ttl action.
watch:[spec...] causes ovs-ofctl to send a ``monitor request''
Nicira extension message to the switch at connection setup time.
This message causes the switch to send information about flow
table changes as they occur. The following comma-separated spec
syntax is available:
!initial
Do not report the switch's initial flow table contents.
!add Do not report newly added flows.
!delete
Do not report deleted flows.
!modify
Do not report modifications to existing flows.
!own Abbreviate changes made to the flow table by ovs-ofctl's
own connection to the switch. (These could only occur
using the ofctl/send command described below under RUN‐
TIME MANAGEMENT COMMANDS.)
!actions
Do not report actions as part of flow updates.
table=number
Limits the monitoring to the table with the given number
between 0 and 254. By default, all tables are monitored.
out_port=port
If set, only flows that output to port are monitored.
The port may be an OpenFlow port number or keyword (e.g.
LOCAL).
field=value
Monitors only flows that have field specified as the
given value. Any syntax valid for matching on dump-flows
may be used.
This command may be useful for debugging switch or controller
implementations. With watch:, it is particularly useful for
observing how a controller updates flow tables.
OpenFlow Switch and Controller Commands
The following commands, like those in the previous section, may be
applied to OpenFlow switches, using any of the connection methods
described in that section. Unlike those commands, these may also be
applied to OpenFlow controllers.
probe target
Sends a single OpenFlow echo-request message to target and waits
for the response. With the -t or --timeout option, this command
can test whether an OpenFlow switch or controller is up and run‐
ning.
ping target [n]
Sends a series of 10 echo request packets to target and times
each reply. The echo request packets consist of an OpenFlow
header plus n bytes (default: 64) of randomly generated payload.
This measures the latency of individual requests.
benchmark target n count
Sends count echo request packets that each consist of an Open‐
Flow header plus n bytes of payload and waits for each response.
Reports the total time required. This is a measure of the maxi‐
mum bandwidth to target for round-trips of n-byte messages.
Other Commands
ofp-parse file
Reads file (or stdin if file is -) as a series of OpenFlow mes‐
sages in the binary format used on an OpenFlow connection, and
prints them to the console. This can be useful for printing
OpenFlow messages captured from a TCP stream.
ofp-parse-pcap file [port...]
Reads file, which must be in the PCAP format used by network
capture tools such as tcpdump or wireshark, extracts all the TCP
streams for OpenFlow connections, and prints the OpenFlow mes‐
sages in those connections in human-readable format on stdout.
OpenFlow connections are distinguished by TCP port number. Non-
OpenFlow packets are ignored. By default, data on TCP ports
6633 and 6653 are considered to be OpenFlow. Specify one or
more port arguments to override the default.
This command cannot usefully print SSL encrypted traffic. It
does not understand IPv6.
Flow Syntax
Some ovs-ofctl commands accept an argument that describes a flow or
flows. Such flow descriptions comprise a series field=value assign‐
ments, separated by commas or white space. (Embedding spaces into a
flow description normally requires quoting to prevent the shell from
breaking the description into multiple arguments.)
Flow descriptions should be in normal form. This means that a flow may
only specify a value for an L3 field if it also specifies a particular
L2 protocol, and that a flow may only specify an L4 field if it also
specifies particular L2 and L3 protocol types. For example, if the L2
protocol type dl_type is wildcarded, then L3 fields nw_src, nw_dst, and
nw_proto must also be wildcarded. Similarly, if dl_type or nw_proto
(the L3 protocol type) is wildcarded, so must be the L4 fields tcp_dst
and tcp_src. ovs-ofctl will warn about flows not in normal form.
The following field assignments describe how a flow matches a packet.
If any of these assignments is omitted from the flow syntax, the field
is treated as a wildcard; thus, if all of them are omitted, the result‐
ing flow matches all packets. The string * may be specified to explic‐
itly mark any of these fields as a wildcard. (* should be quoted to
protect it from shell expansion.)
in_port=port
Matches OpenFlow port port, which may be an OpenFlow port number
or keyword (e.g. LOCAL). ovs-ofctl show.
(The resubmit action can search OpenFlow flow tables with arbi‐
trary in_port values, so flows that match port numbers that do
not exist from an OpenFlow perspective can still potentially be
matched.)
dl_vlan=vlan
Matches IEEE 802.1q Virtual LAN tag vlan. Specify 0xffff as
vlan to match packets that are not tagged with a Virtual LAN;
otherwise, specify a number between 0 and 4095, inclusive, as
the 12-bit VLAN ID to match.
dl_vlan_pcp=priority
Matches IEEE 802.1q Priority Code Point (PCP) priority, which is
specified as a value between 0 and 7, inclusive. A higher value
indicates a higher frame priority level.
dl_src=xx:xx:xx:xx:xx:xx
dl_dst=xx:xx:xx:xx:xx:xx
Matches an Ethernet source (or destination) address specified as
6 pairs of hexadecimal digits delimited by colons (e.g.
00:0A:E4:25:6B:B0).
dl_src=xx:xx:xx:xx:xx:xx/xx:xx:xx:xx:xx:xx
dl_dst=xx:xx:xx:xx:xx:xx/xx:xx:xx:xx:xx:xx
Matches an Ethernet destination address specified as 6 pairs of
hexadecimal digits delimited by colons (e.g. 00:0A:E4:25:6B:B0),
with a wildcard mask following the slash. Open vSwitch 1.8 and
later support arbitrary masks for source and/or destination.
Earlier versions only support masking the destination with the
following masks:
01:00:00:00:00:00
Match only the multicast bit. Thus,
dl_dst=01:00:00:00:00:00/01:00:00:00:00:00 matches all
multicast (including broadcast) Ethernet packets, and
dl_dst=00:00:00:00:00:00/01:00:00:00:00:00 matches all
unicast Ethernet packets.
fe:ff:ff:ff:ff:ff
Match all bits except the multicast bit. This is proba‐
bly not useful.
ff:ff:ff:ff:ff:ff
Exact match (equivalent to omitting the mask).
00:00:00:00:00:00
Wildcard all bits (equivalent to dl_dst=*.)
dl_type=ethertype
Matches Ethernet protocol type ethertype, which is specified as
an integer between 0 and 65535, inclusive, either in decimal or
as a hexadecimal number prefixed by 0x (e.g. 0x0806 to match ARP
packets).
nw_src=ip[/netmask]
nw_dst=ip[/netmask]
When dl_type is 0x0800 (possibly via shorthand, e.g. ip or tcp),
matches IPv4 source (or destination) address ip, which may be
specified as an IP address or host name (e.g. 192.168.1.1 or
www.example.com). The optional netmask allows restricting a
match to an IPv4 address prefix. The netmask may be specified
as a dotted quad (e.g. 192.168.1.0/255.255.255.0) or as a CIDR
block (e.g. 192.168.1.0/24). Open vSwitch 1.8 and later support
arbitrary dotted quad masks; earlier versions support only CIDR
masks, that is, the dotted quads that are equivalent to some
CIDR block.
When dl_type=0x0806 or arp is specified, matches the ar_spa or
ar_tpa field, respectively, in ARP packets for IPv4 and Ether‐
net.
When dl_type=0x8035 or rarp is specified, matches the ar_spa or
ar_tpa field, respectively, in RARP packets for IPv4 and Ether‐
net.
When dl_type is wildcarded or set to a value other than 0x0800,
0x0806, or 0x8035, the values of nw_src and nw_dst are ignored
(see Flow Syntax above).
nw_proto=proto
ip_proto=proto
When ip or dl_type=0x0800 is specified, matches IP protocol type
proto, which is specified as a decimal number between 0 and 255,
inclusive (e.g. 1 to match ICMP packets or 6 to match TCP pack‐
ets).
When ipv6 or dl_type=0x86dd is specified, matches IPv6 header
type proto, which is specified as a decimal number between 0 and
255, inclusive (e.g. 58 to match ICMPv6 packets or 6 to match
TCP). The header type is the terminal header as described in
the DESIGN document.
When arp or dl_type=0x0806 is specified, matches the lower 8
bits of the ARP opcode. ARP opcodes greater than 255 are
treated as 0.
When rarp or dl_type=0x8035 is specified, matches the lower 8
bits of the ARP opcode. ARP opcodes greater than 255 are
treated as 0.
When dl_type is wildcarded or set to a value other than 0x0800,
0x0806, 0x8035 or 0x86dd, the value of nw_proto is ignored (see
Flow Syntax above).
nw_tos=tos
Matches IP ToS/DSCP or IPv6 traffic class field tos, which is
specified as a decimal number between 0 and 255, inclusive.
Note that the two lower reserved bits are ignored for matching
purposes.
When dl_type is wildcarded or set to a value other than 0x0800
or 0x86dd, the value of nw_tos is ignored (see Flow Syntax
above).
ip_dscp=dscp
Matches IP ToS/DSCP or IPv6 traffic class field dscp, which is
specified as a decimal number between 0 and 63, inclusive.
When dl_type is wildcarded or set to a value other than 0x0800
or 0x86dd, the value of ip_dscp is ignored (see Flow Syntax
above).
nw_ecn=ecn
ip_ecn=ecn
Matches ecn bits in IP ToS or IPv6 traffic class fields, which
is specified as a decimal number between 0 and 3, inclusive.
When dl_type is wildcarded or set to a value other than 0x0800
or 0x86dd, the value of nw_ecn is ignored (see Flow Syntax
above).
nw_ttl=ttl
Matches IP TTL or IPv6 hop limit value ttl, which is specified
as a decimal number between 0 and 255, inclusive.
When dl_type is wildcarded or set to a value other than 0x0800
or 0x86dd, the value of nw_ttl is ignored (see Flow Syntax
above).
tcp_src=port
tcp_dst=port
udp_src=port
udp_dst=port
sctp_src=port
sctp_dst=port
Matches a TCP, UDP, or SCTP source or destination port port,
which is specified as a decimal number between 0 and 65535,
inclusive.
When dl_type and nw_proto are wildcarded or set to values that
do not indicate an appropriate protocol, the values of these
settings are ignored (see Flow Syntax above).
tcp_src=port/mask
tcp_dst=port/mask
udp_src=port/mask
udp_dst=port/mask
sctp_src=port/mask
sctp_dst=port/mask
Bitwise match on TCP (or UDP or SCTP) source or destination
port. The port and mask are 16-bit numbers written in decimal
or in hexadecimal prefixed by 0x. Each 1-bit in mask requires
that the corresponding bit in port must match. Each 0-bit in
mask causes the corresponding bit to be ignored.
Bitwise matches on transport ports are rarely useful in isola‐
tion, but a group of them can be used to reduce the number of
flows required to match on a range of transport ports. For
example, suppose that the goal is to match TCP source ports 1000
to 1999, inclusive. One way is to insert 1000 flows, each of
which matches on a single source port. Another way is to look
at the binary representations of 1000 and 1999, as follows:
01111101000
11111001111
and then to transform those into a series of bitwise matches
that accomplish the same results:
01111101xxx
0111111xxxx
10xxxxxxxxx
110xxxxxxxx
1110xxxxxxx
11110xxxxxx
1111100xxxx
which become the following when written in the syntax required
by ovs-ofctl:
tcp,tcp_src=0x03e8/0xfff8
tcp,tcp_src=0x03f0/0xfff0
tcp,tcp_src=0x0400/0xfe00
tcp,tcp_src=0x0600/0xff00
tcp,tcp_src=0x0700/0xff80
tcp,tcp_src=0x0780/0xffc0
tcp,tcp_src=0x07c0/0xfff0
Only Open vSwitch 1.6 and later supports bitwise matching on
transport ports.
Like the exact-match forms described above, the bitwise match
forms apply only when dl_type and nw_proto specify TCP or UDP or
SCTP.
tp_src=port
tp_dst=port
These are deprecated generic forms of L4 port matches. In new
code, please use the TCP-, UDP-, or SCTP-specific forms
described above.
tcp_flags=flags/mask
tcp_flags=[+flag...][-flag...]
Bitwise match on TCP flags. The flags and mask are 16-bit num‐
bers written in decimal or in hexadecimal prefixed by 0x. Each
1-bit in mask requires that the corresponding bit in flags must
match. Each 0-bit in mask causes the corresponding bit to be
ignored.
Alternatively, the flags can be specified by their symbolic
names (listed below), each preceded by either + for a flag that
must be set, or - for a flag that must be unset, without any
other delimiters between the flags. Flags not mentioned are
wildcarded. For example, tcp,tcp_flags=+syn-ack matches TCP
SYNs that are not ACKs.
TCP protocol currently defines 9 flag bits, and additional 3
bits are reserved (must be transmitted as zero), see RFCs 793,
3168, and 3540. The flag bits are, numbering from the least
significant bit:
0: fin No more data from sender.
1: syn Synchronize sequence numbers.
2: rst Reset the connection.
3: psh Push function.
4: ack Acknowledgement field significant.
5: urg Urgent pointer field significant.
6: ece ECN Echo.
7: cwr Congestion Windows Reduced.
8: ns Nonce Sum.
9-11: Reserved.
12-15: Not matchable, must be zero.
icmp_type=type
icmp_code=code
When dl_type and nw_proto specify ICMP or ICMPv6, type matches
the ICMP type and code matches the ICMP code. Each is specified
as a decimal number between 0 and 255, inclusive.
When dl_type and nw_proto take other values, the values of these
settings are ignored (see Flow Syntax above).
table=number
For flow dump commands, limits the flows dumped to those in the
table with the given number between 0 and 254. If not specified
(or if 255 is specified as number), then flows in all tables are
dumped.
For flow table modification commands, behavior varies based on
the OpenFlow version used to connect to the switch:
OpenFlow 1.0
OpenFlow 1.0 does not support table for modifying flows.
ovs-ofctl will exit with an error if table (other than
table=255) is specified for a switch that only supports
OpenFlow 1.0.
In OpenFlow 1.0, the switch chooses the table into which
to insert a new flow. The Open vSwitch software switch
always chooses table 0. Other Open vSwitch datapaths and
other OpenFlow implementations may choose different
tables.
The OpenFlow 1.0 behavior in Open vSwitch for modifying
or removing flows depends on whether --strict is used.
Without --strict, the command applies to matching flows
in all tables. With --strict, the command will operate
on any single matching flow in any table; it will do
nothing if there are matches in more than one table.
(The distinction between these behaviors only matters if
non-OpenFlow 1.0 commands were also used, because Open‐
Flow 1.0 alone cannot add flows with the same matching
criteria to multiple tables.)
OpenFlow 1.0 with table_id extension
Open vSwitch implements an OpenFlow extension that allows
the controller to specify the table on which to operate.
ovs-ofctl automatically enables the extension when table
is specified and OpenFlow 1.0 is used. ovs-ofctl auto‐
matically detects whether the switch supports the exten‐
sion. As of this writing, this extension is only known
to be implemented by Open vSwitch.
With this extension, ovs-ofctl operates on the requested
table when table is specified, and acts as described for
OpenFlow 1.0 above when no table is specified (or for ta‐
ble=255).
OpenFlow 1.1
OpenFlow 1.1 requires flow table modification commands to
specify a table. When table is not specified (or ta‐
ble=255 is specified), ovs-ofctl defaults to table 0.
OpenFlow 1.2 and later
OpenFlow 1.2 and later allow flow deletion commands, but
not other flow table modification commands, to operate on
all flow tables, with the behavior described above for
OpenFlow 1.0.
metadata=value[/mask]
Matches value either exactly or with optional mask in the meta‐
data field. value and mask are 64-bit integers, by default in
decimal (use a 0x prefix to specify hexadecimal). Arbitrary mask
values are allowed: a 1-bit in mask indicates that the corre‐
sponding bit in value must match exactly, and a 0-bit wildcards
that bit. Matching on metadata was added in Open vSwitch 1.8.
The following shorthand notations are also available:
ip Same as dl_type=0x0800.
ipv6 Same as dl_type=0x86dd.
icmp Same as dl_type=0x0800,nw_proto=1.
icmp6 Same as dl_type=0x86dd,nw_proto=58.
tcp Same as dl_type=0x0800,nw_proto=6.
tcp6 Same as dl_type=0x86dd,nw_proto=6.
udp Same as dl_type=0x0800,nw_proto=17.
udp6 Same as dl_type=0x86dd,nw_proto=17.
sctp Same as dl_type=0x0800,nw_proto=132.
sctp6 Same as dl_type=0x86dd,nw_proto=132.
arp Same as dl_type=0x0806.
rarp Same as dl_type=0x8035.
mpls Same as dl_type=0x8847.
mplsm Same as dl_type=0x8848.
The following field assignments require support for the NXM (Nicira
Extended Match) extension to OpenFlow. When one of these is specified,
ovs-ofctl will automatically attempt to negotiate use of this exten‐
sion. If the switch does not support NXM, then ovs-ofctl will report a
fatal error.
vlan_tci=tci[/mask]
Matches modified VLAN TCI tci. If mask is omitted, tci is the
exact VLAN TCI to match; if mask is specified, then a 1-bit in
mask indicates that the corresponding bit in tci must match
exactly, and a 0-bit wildcards that bit. Both tci and mask are
16-bit values that are decimal by default; use a 0x prefix to
specify them in hexadecimal.
The value that vlan_tci matches against is 0 for a packet that
has no 802.1Q header. Otherwise, it is the TCI value from the
802.1Q header with the CFI bit (with value 0x1000) forced to 1.
Examples:
vlan_tci=0
Match only packets without an 802.1Q header.
vlan_tci=0xf123
Match packets tagged with priority 7 in VLAN 0x123.
vlan_tci=0x1123/0x1fff
Match packets tagged with VLAN 0x123 (and any priority).
vlan_tci=0x5000/0xf000
Match packets tagged with priority 2 (in any VLAN).
vlan_tci=0/0xfff
Match packets with no 802.1Q header or tagged with VLAN 0
(and any priority).
vlan_tci=0x5000/0xe000
Match packets with no 802.1Q header or tagged with prior‐
ity 2 (in any VLAN).
vlan_tci=0/0xefff
Match packets with no 802.1Q header or tagged with VLAN 0
and priority 0.
Some of these matching possibilities can also be achieved with
dl_vlan and dl_vlan_pcp.
ip_frag=frag_type
When dl_type specifies IP or IPv6, frag_type specifies what kind
of IP fragments or non-fragments to match. The following values
of frag_type are supported:
no Matches only non-fragmented packets.
yes Matches all fragments.
first Matches only fragments with offset 0.
later Matches only fragments with nonzero offset.
not_later
Matches non-fragmented packets and fragments with zero
offset.
The ip_frag match type is likely to be most useful in nx-match
mode. See the description of the set-frags command, above, for
more details.
arp_spa=ip[/netmask]
arp_tpa=ip[/netmask]
When dl_type specifies either ARP or RARP, arp_spa and arp_tpa
match the source and target IPv4 address, respectively. An
address may be specified as an IP address or host name (e.g.
192.168.1.1 or www.example.com). The optional netmask allows
restricting a match to an IPv4 address prefix. The netmask may
be specified as a dotted quad (e.g. 192.168.1.0/255.255.255.0)
or as a CIDR block (e.g. 192.168.1.0/24).
arp_sha=xx:xx:xx:xx:xx:xx
arp_tha=xx:xx:xx:xx:xx:xx
When dl_type specifies either ARP or RARP, arp_sha and arp_tha
match the source and target hardware address, respectively. An
address is specified as 6 pairs of hexadecimal digits delimited
by colons (e.g. 00:0A:E4:25:6B:B0).
arp_sha=xx:xx:xx:xx:xx:xx/xx:xx:xx:xx:xx:xx
arp_tha=xx:xx:xx:xx:xx:xx/xx:xx:xx:xx:xx:xx
When dl_type specifies either ARP or RARP, arp_sha and arp_tha
match the source and target hardware address, respectively. An
address is specified as 6 pairs of hexadecimal digits delimited
by colons (e.g. 00:0A:E4:25:6B:B0), with a wildcard mask follow‐
ing the slash.
ipv6_src=ipv6[/netmask]
ipv6_dst=ipv6[/netmask]
When dl_type is 0x86dd (possibly via shorthand, e.g., ipv6 or
tcp6), matches IPv6 source (or destination) address ipv6, which
may be specified as defined in RFC 2373. The preferred format
is x:x:x:x:x:x:x:x, where x are the hexadecimal values of the
eight 16-bit pieces of the address. A single instance of :: may
be used to indicate multiple groups of 16-bits of zeros. The
optional netmask allows restricting a match to an IPv6 address
prefix. A netmask is specified as an IPv6 address (e.g.
2001:db8:3c4d:1::/ffff:ffff:ffff:ffff::) or a CIDR block (e.g.
2001:db8:3c4d:1::/64). Open vSwitch 1.8 and later support arbi‐
trary masks; earlier versions support only CIDR masks, that is,
CIDR block and IPv6 addresses that are equivalent to CIDR
blocks.
ipv6_label=label
When dl_type is 0x86dd (possibly via shorthand, e.g., ipv6 or
tcp6), matches IPv6 flow label label.
nd_target=ipv6[/netmask]
When dl_type, nw_proto, and icmp_type specify IPv6 Neighbor Dis‐
covery (ICMPv6 type 135 or 136), matches the target address
ipv6. ipv6 is in the same format described earlier for the
ipv6_src and ipv6_dst fields.
nd_sll=xx:xx:xx:xx:xx:xx
When dl_type, nw_proto, and icmp_type specify IPv6 Neighbor
Solicitation (ICMPv6 type 135), matches the source link-layer
address option. An address is specified as 6 pairs of hexadeci‐
mal digits delimited by colons.
nd_tll=xx:xx:xx:xx:xx:xx
When dl_type, nw_proto, and icmp_type specify IPv6 Neighbor
Advertisement (ICMPv6 type 136), matches the target link-layer
address option. An address is specified as 6 pairs of hexadeci‐
mal digits delimited by colons.
mpls_bos=bos
When dl_type is 0x8847 or 0x8848 (possibly via shorthand e.g.,
mpls or mplsm), matches the bottom-of-stack bit of the outer-
most MPLS label stack entry. Valid values are 0 and 1.
If 1 then for a packet with a well-formed MPLS label stack the
bottom-of-stack bit indicates that the outer label stack entry
is also the inner-most label stack entry and thus that is that
there is only one label stack entry present. Conversely, if 0
then for a packet with a well-formed MPLS label stack the bot‐
tom-of-stack bit indicates that the outer label stack entry is
not the inner-most label stack entry and thus there is more than
one label stack entry present.
mpls_label=label
When dl_type is 0x8847 or 0x8848 (possibly via shorthand e.g.,
mpls or mplsm), matches the label of the outer MPLS label stack
entry. The label is a 20-bit value that is decimal by default;
use a 0x prefix to specify them in hexadecimal.
mpls_tc=tc
When dl_type is 0x8847 or 0x8848 (possibly via shorthand e.g.,
mpls or mplsm), matches the traffic-class of the outer MPLS
label stack entry. Valid values are between 0 (lowest) and 7
(highest).
tun_id=tunnel-id[/mask]
tunnel_id=tunnel-id[/mask]
Matches tunnel identifier tunnel-id. Only packets that arrive
over a tunnel that carries a key (e.g. GRE with the RFC 2890 key
extension and a nonzero key value) will have a nonzero tunnel
ID. If mask is omitted, tunnel-id is the exact tunnel ID to
match; if mask is specified, then a 1-bit in mask indicates that
the corresponding bit in tunnel-id must match exactly, and a
0-bit wildcards that bit.
tun_flags=flags
Matches flags indicating various aspects of the tunnel encapsu‐
lation. Currently, there is only one flag defined:
oam: The tunnel protocol indicated that this is an OAM control
packet.
Flags can be prefixed by + or - to indicate that the flag should
be matched as either present or not present, respectively. In
addition, flags can be specified without a prefix and separated
by | to indicate an exact match.
Note that it is possible for newer version of Open vSwitch to
introduce additional flags with varying meaning. It is therefore
not recommended to use an exact match on this field since the
behavior of these new flags is unknown and should be ignored.
For non-tunneled packets, the value is 0.
This field was introduced in Open vSwitch 2.5.
tun_src=ip[/netmask]
tun_dst=ip[/netmask]
Matches tunnel IPv4 source (or destination) address ip. Only
packets that arrive over a tunnel will have nonzero tunnel
addresses. The address may be specified as an IP address or
host name (e.g. 192.168.1.1 or www.example.com). The optional
netmask allows restricting a match to a masked IPv4 address.
The netmask may be specified as a dotted quad (e.g.
192.168.1.0/255.255.255.0) or as a CIDR block (e.g.
192.168.1.0/24).
tun_gbp_id=value[/mask]
tun_gbp_flags=value[/mask]
Matches the group policy identifier and flags in the VXLAN
header. Only packets that arrive over a VXLAN tunnel with the
"gbp" extension enabled can have this field set. The fields may
also be referred to by NXM_NX_TUN_GBP_ID[] (16 bits) and
NXM_NX_TUN_GBP_FLAGS[] (8 bits) in the context of field manipu‐
lation actions. If these fields are set and the packet matched
by the flow is encapsulated in a VXLAN-GBP tunnel, then the pol‐
icy identifier and flags are transmitted to the destination
VXLAN tunnel endpoint.
The tun_gbp_flags field has the following format:
+-+-+-+-+-+-+-+-+
|-|D|-|-|A|-|-|-|
+-+-+-+-+-+-+-+-+
D := Don't Learn bit. When set, this bit indicates that the
egress tunnel endpoint MUST NOT learn the source address of
the encapsulated frame.
A := Indicates that the group policy has already been applied
to this packet. Policies MUST NOT be applied by devices when
the A bit is set.
For more information, please see the corresponding IETF draft:
https://tools.ietf.org/html/draft-smith-vxlan-group-policy
tun_metadataidx[=value[/mask]]
Matches value either exactly or with optional mask in tunnel
metadata field number idx (numbered from 0 to 63). The act of
specifying a field implies a match on the existence of that
field in the packet in addition to the masked value. As a short‐
hand, it is possible to specify only the field name to simply
match on an option being present.
Tunnel metadata fields can be dynamically assigned onto the data
contained in the options of Geneve packets using the commands
described in the section OpenFlow Switch Geneve Option Table
Commands. Once assigned, the length of the field is variable
according to the size of the option. Before updating a mapping
in the option table, flows with references to it should be
removed, otherwise the result is non-deterministic.
These fields were introduced in Open vSwitch 2.5.
regidx=value[/mask]
Matches value either exactly or with optional mask in register
number idx. The valid range of idx depends on the switch.
value and mask are 32-bit integers, by default in decimal (use a
0x prefix to specify hexadecimal). Arbitrary mask values are
allowed: a 1-bit in mask indicates that the corresponding bit in
value must match exactly, and a 0-bit wildcards that bit.
When a packet enters an OpenFlow switch, all of the registers
are set to 0. Only explicit actions change register values.
xregidx=value[/mask]
Matches value either exactly or with optional mask in 64-bit
``extended register'' number idx. Each of the 64-bit extended
registers overlays two of the 32-bit registers: xreg0 overlays
reg0 and reg1, with reg0 supplying the most-significant bits of
xreg0 and reg1 the least-significant. xreg1 similarly overlays
reg2 and reg3, and so on.
These fields were added in Open vSwitch 2.3 to conform with the
OpenFlow 1.5 specification. OpenFlow 1.5 calls these fields
just the ``packet registers,'' but Open vSwitch already had
32-bit registers by that name, which is why Open vSwitch refers
to the standard registers as ``extended registers''.
pkt_mark=value[/mask]
Matches packet metadata mark value either exactly or with
optional mask. The mark is associated data that may be passed
into other system components in order to facilitate interaction
between subsystems. On Linux this corresponds to the skb mark
but the exact implementation is platform-dependent.
actset_output=port
Matches the output port currently in the OpenFlow action set,
where port may be an OpenFlow port number or keyword (e.g.
LOCAL). If there is no output port in the OpenFlow action set,
or if the output port will be ignored (e.g. because there is an
output group in the OpenFlow action set), then the value will be
UNSET.
This field was introduced in Open vSwitch 2.4 to conform with
the OpenFlow 1.5 specification.
conj_id=value
Matches the given 32-bit value against the conjunction ID. This
is used only with the conjunction action (see below).
This field was introduced in Open vSwitch 2.4.
Defining IPv6 flows (those with dl_type equal to 0x86dd) requires sup‐
port for NXM. The following shorthand notations are available for
IPv6-related flows:
ipv6 Same as dl_type=0x86dd.
tcp6 Same as dl_type=0x86dd,nw_proto=6.
udp6 Same as dl_type=0x86dd,nw_proto=17.
sctp6 Same as dl_type=0x86dd,nw_proto=132.
icmp6 Same as dl_type=0x86dd,nw_proto=58.
Finally, field assignments to duration, n_packets, or n_bytes are
ignored to allow output from the dump-flows command to be used as input
for other commands that parse flows.
The add-flow, add-flows, and mod-flows commands require an additional
field, which must be the final field specified:
actions=[action][,action...]
Specifies a comma-separated list of actions to take on a packet
when the flow entry matches. If no action is specified, then
packets matching the flow are dropped. The following forms of
action are supported:
port
output:port
Outputs the packet to OpenFlow port number port. If port
is the packet's input port, the packet is not output.
output:src[start..end]
Outputs the packet to the OpenFlow port number read from
src, which must be an NXM field as described above. For
example, output:NXM_NX_REG0[16..31] outputs to the Open‐
Flow port number written in the upper half of register 0.
If the port number is the packet's input port, the packet
is not output.
This form of output was added in Open vSwitch 1.3.0.
This form of output uses an OpenFlow extension that is
not supported by standard OpenFlow switches.
group:group_id
Outputs the packet to the OpenFlow group group_id. Group
tables are only supported in OpenFlow 1.1+. See Group
Syntax for more details.
normal Subjects the packet to the device's normal L2/L3 process‐
ing. (This action is not implemented by all OpenFlow
switches.)
flood Outputs the packet on all switch physical ports other
than the port on which it was received and any ports on
which flooding is disabled (typically, these would be
ports disabled by the IEEE 802.1D spanning tree proto‐
col).
all Outputs the packet on all switch physical ports other
than the port on which it was received.
local Outputs the packet on the ``local port,'' which corre‐
sponds to the network device that has the same name as
the bridge.
in_port
Outputs the packet on the port from which it was
received.
controller(key=value...)
Sends the packet to the OpenFlow controller as a ``packet
in'' message. The supported key-value pairs are:
max_len=nbytes
Limit to nbytes the number of bytes of the packet
to send to the controller. By default the entire
packet is sent.
reason=reason
Specify reason as the reason for sending the mes‐
sage in the ``packet in'' message. The supported
reasons are action (the default), no_match, and
invalid_ttl.
id=controller-id
Specify controller-id, a 16-bit integer, as the
connection ID of the OpenFlow controller or con‐
trollers to which the ``packet in'' message should
be sent. The default is zero. Zero is also the
default connection ID for each controller connec‐
tion, and a given controller connection will only
have a nonzero connection ID if its controller
uses the NXT_SET_CONTROLLER_ID Nicira extension to
OpenFlow.
Any reason other than action and any nonzero controller-
id uses a Nicira vendor extension that, as of this writ‐
ing, is only known to be implemented by Open vSwitch
(version 1.6 or later).
controller
controller[:nbytes]
Shorthand for controller() or controller(max_len=nbytes),
respectively.
enqueue(port,queue)
Enqueues the packet on the specified queue within port
port, which must be an OpenFlow port number or keyword
(e.g. LOCAL). The number of supported queues depends on
the switch; some OpenFlow implementations do not support
queuing at all.
drop Discards the packet, so no further processing or forward‐
ing takes place. If a drop action is used, no other
actions may be specified.
mod_vlan_vid:vlan_vid
Modifies the VLAN id on a packet. The VLAN tag is added
or modified as necessary to match the value specified.
If the VLAN tag is added, a priority of zero is used (see
the mod_vlan_pcp action to set this).
mod_vlan_pcp:vlan_pcp
Modifies the VLAN priority on a packet. The VLAN tag is
added or modified as necessary to match the value speci‐
fied. Valid values are between 0 (lowest) and 7 (high‐
est). If the VLAN tag is added, a vid of zero is used
(see the mod_vlan_vid action to set this).
strip_vlan
Strips the VLAN tag from a packet if it is present.
push_vlan:ethertype
Push a new VLAN tag onto the packet. Ethertype is used
as the Ethertype for the tag. Only ethertype 0x8100
should be used. (0x88a8 which the spec allows isn't sup‐
ported at the moment.) A priority of zero and the tag of
zero are used for the new tag.
push_mpls:ethertype
Changes the packet's Ethertype to ethertype, which must
be either 0x8847 or 0x8848, and pushes an MPLS LSE.
If the packet does not already contain any MPLS labels
then an initial label stack entry is pushed. The label
stack entry's label is 2 if the packet contains IPv6 and
0 otherwise, its default traffic control value is the low
3 bits of the packet's DSCP value (0 if the packet is not
IP), and its TTL is copied from the IP TTL (64 if the
packet is not IP).
If the packet does already contain an MPLS label, pushes
a new outermost label as a copy of the existing outermost
label.
A limitation of the implementation is that processing of
actions will stop if push_mpls follows another push_mpls
unless there is a pop_mpls in between.
pop_mpls:ethertype
Strips the outermost MPLS label stack entry. Currently
the implementation restricts ethertype to a non-MPLS
Ethertype and thus pop_mpls should only be applied to
packets with an MPLS label stack depth of one. A further
limitation is that processing of actions will stop if
pop_mpls follows another pop_mpls unless there is a
push_mpls in between.
mod_dl_src:mac
Sets the source Ethernet address to mac.
mod_dl_dst:mac
Sets the destination Ethernet address to mac.
mod_nw_src:ip
Sets the IPv4 source address to ip.
mod_nw_dst:ip
Sets the IPv4 destination address to ip.
mod_tp_src:port
Sets the TCP or UDP or SCTP source port to port.
mod_tp_dst:port
Sets the TCP or UDP or SCTP destination port to port.
mod_nw_tos:tos
Sets the DSCP bits in the IPv4 ToS/DSCP or IPv6 traffic
class field to tos, which must be a multiple of 4 between
0 and 255. This action does not modify the two least
significant bits of the ToS field (the ECN bits).
mod_nw_ecn:ecn
Sets the ECN bits in the IPv4 ToS or IPv6 traffic class
field to ecn, which must be a value between 0 and 3,
inclusive. This action does not modify the six most sig‐
nificant bits of the field (the DSCP bits).
Requires OpenFlow 1.1 or later.
mod_nw_ttl:ttl
Sets the IPv4 TTL or IPv6 hop limit field to ttl, which
is specified as a decimal number between 0 and 255,
inclusive. Switch behavior when setting ttl to zero is
not well specified, though.
Requires OpenFlow 1.1 or later.
The following actions are Nicira vendor extensions that, as of
this writing, are only known to be implemented by Open vSwitch:
resubmit:port
resubmit([port],[table])
Re-searches this OpenFlow flow table (or the table whose
number is specified by table) with the in_port field
replaced by port (if port is specified) and executes the
actions found, if any, in addition to any other actions
in this flow entry.
Recursive resubmit actions are obeyed up to an implemen‐
tation-defined maximum depth. Open vSwitch 1.0.1 and
earlier did not support recursion; Open vSwitch before
1.2.90 did not support table.
set_tunnel:id
set_tunnel64:id
If outputting to a port that encapsulates the packet in a
tunnel and supports an identifier (such as GRE), sets the
identifier to id. If the set_tunnel form is used and id
fits in 32 bits, then this uses an action extension that
is supported by Open vSwitch 1.0 and later. Otherwise,
if id is a 64-bit value, it requires Open vSwitch 1.1 or
later.
set_queue:queue
Sets the queue that should be used to queue when packets
are output. The number of supported queues depends on
the switch; some OpenFlow implementations do not support
queuing at all.
pop_queue
Restores the queue to the value it was before any
set_queue actions were applied.
dec_ttl
dec_ttl[(id1,id2)]
Decrement TTL of IPv4 packet or hop limit of IPv6 packet.
If the TTL or hop limit is initially zero or decrementing
would make it so, no decrement occurs, as packets reach‐
ing TTL zero must be rejected. Instead, a ``packet-in''
message with reason code OFPR_INVALID_TTL is sent to each
connected controller that has enabled receiving them, if
any. Processing the current set of actions then stops.
However, if the current set of actions was reached
through ``resubmit'' then remaining actions in outer lev‐
els resume processing. This action also optionally sup‐
ports the ability to specify a list of valid controller
ids. Each of controllers in the list will receive the
``packet_in'' message only if they have registered to
receive the invalid ttl packets. If controller ids are
not specified, the ``packet_in'' message will be sent
only to the controllers having controller id zero which
have registered for the invalid ttl packets.
set_mpls_label:label
Set the label of the outer MPLS label stack entry of a
packet. label should be a 20-bit value that is decimal
by default; use a 0x prefix to specify them in hexadeci‐
mal.
set_mpls_tc:tc
Set the traffic-class of the outer MPLS label stack entry
of a packet. tc should be a in the range 0 to 7 inclu‐
sive.
set_mpls_ttl:ttl
Set the TTL of the outer MPLS label stack entry of a
packet. ttl should be in the range 0 to 255 inclusive.
dec_mpls_ttl
Decrement TTL of the outer MPLS label stack entry of a
packet. If the TTL is initially zero or decrementing
would make it so, no decrement occurs. Instead, a
``packet-in'' message with reason code OFPR_INVALID_TTL
is sent to the main controller (id zero), if it has
enabled receiving them. Processing the current set of
actions then stops. However, if the current set of
actions was reached through ``resubmit'' then remaining
actions in outer levels resume processing.
note:[hh]...
Does nothing at all. Any number of bytes represented as
hex digits hh may be included. Pairs of hex digits may
be separated by periods for readability. The note
action's format doesn't include an exact length for its
payload, so the provided bytes will be padded on the
right by enough bytes with value 0 to make the total num‐
ber 6 more than a multiple of 8.
move:src[start..end]->gt;>gt;dst[start..end]
Copies the named bits from field src to field dst. src
and dst must be NXM field names as defined in
nicira-ext.h, e.g. NXM_OF_UDP_SRC or NXM_NX_REG0. Each
start and end pair, which are inclusive, must specify the
same number of bits and must fit within its respective
field. Shorthands for [start..end] exist: use [bit] to
specify a single bit or [] to specify an entire field.
Examples: move:NXM_NX_REG0[0..5]->gt;>gt;NXM_NX_REG1[26..31]
copies the six bits numbered 0 through 5, inclusive, in
register 0 into bits 26 through 31, inclusive;
move:NXM_NX_REG0[0..15]->gt;>gt;NXM_OF_VLAN_TCI[] copies the
least significant 16 bits of register 0 into the VLAN TCI
field.
In OpenFlow 1.0 through 1.4, move ordinarily uses an Open
vSwitch extension to OpenFlow. In OpenFlow 1.5, move
uses the OpenFlow 1.5 standard copy_field action. The
ONF has also made copy_field available as an extension to
OpenFlow 1.3. Open vSwitch 2.4 and later understands
this extension and uses it if a controller uses it, but
for backward compatibility with older versions of Open
vSwitch, ovs-ofctl does not use it.
set_field:value[/mask]->gt;>gt;dst
load:value->gt;>gt;dst[start..end]
Loads a literal value into a field or part of a field.
With set_field, value and the optional mask are given in
the customary syntax for field dst, which is expressed as
a field name. For example,
set_field:00:11:22:33:44:55->gt;>gt;eth_src sets the Ethernet
source address to 00:11:22:33:44:55. With load, value
must be an integer value (in decimal or prefixed by 0x
for hexadecimal) and dst is the NXM or OXM name for the
field. For example,
load:0x001122334455->gt;>gt;OXM_OF_ETH_DST[] has the same effect
as the prior set_field example.
The two forms exist for historical reasons. Open vSwitch
1.1 introduced NXAST_REG_LOAD as a Nicira extension to
OpenFlow 1.0 and used load to express it. Later, Open‐
Flow 1.2 introduced a standard OFPAT_SET_FIELD action
that was restricted to loading entire fields, so Open
vSwitch added the form set_field with this restriction.
OpenFlow 1.5 extended OFPAT_SET_FIELD to the point that
it became a superset of NXAST_REG_LOAD. Open vSwitch
translates either syntax as necessary for the OpenFlow
version in use: in OpenFlow 1.0 and 1.1, NXAST_REG_LOAD;
in OpenFlow 1.2, 1.3, and 1.4, NXAST_REG_LOAD for load or
for loading a subfield, OFPAT_SET_FIELD otherwise; and
OpenFlow 1.5 and later, OFPAT_SET_FIELD.
push:src[start..end]
Pushes start to end bits inclusive, in fields on top of
the stack.
Example: push:NXM_NX_REG2[0..5] push the value stored in
register 2 bits 0 through 5, inclusive, on to the inter‐
nal stack.
pop:dst[start..end]
Pops from the top of the stack, retrieves the start to
end bits inclusive, from the value popped and store them
into the corresponding bits in dst.
Example: pop:NXM_NX_REG2[0..5] pops the value from top of
the stack. Set register 2 bits 0 through 5, inclusive,
based on bits 0 through 5 from the value just popped.
multipath(fields, basis, algorithm, n_links, arg,
dst[start..end])
Hashes fields using basis as a universal hash parameter,
then the applies multipath link selection algorithm (with
parameter arg) to choose one of n_links output links num‐
bered 0 through n_links minus 1, and stores the link into
dst[start..end], which must be an NXM field as described
above.
fields must be one of the following:
eth_src
Hashes Ethernet source address only.
symmetric_l4
Hashes Ethernet source, destination, and type,
VLAN ID, IPv4/IPv6 source, destination, and proto‐
col, and TCP or SCTP (but not UDP) ports. The
hash is computed so that pairs of corresponding
flows in each direction hash to the same value, in
environments where L2 paths are the same in each
direction. UDP ports are not included in the hash
to support protocols such as VXLAN that use asym‐
metric ports in each direction.
symmetric_l3l4
Hashes IPv4/IPv6 source, destination, and proto‐
col, and TCP or SCTP (but not UDP) ports. Like
symmetric_l4, this is a symmetric hash, but by
excluding L2 headers it is more effective in envi‐
ronments with asymmetric L2 paths (e.g. paths
involving VRRP IP addresses on a router). Not an
effective hash function for protocols other than
IPv4 and IPv6, which hash to a constant zero.
symmetric_l3l4+udp
Like symmetric_l3l4+udp, but UDP ports are
included in the hash. This is a more effective
hash when asymmetric UDP protocols such as VXLAN
are not a consideration.
algorithm must be one of modulo_n, hash_threshold, hrw,
and iter_hash. Only the iter_hash algorithm uses arg.
Refer to nicira-ext.h for more details.
bundle(fields, basis, algorithm, slave_type, slaves:[s1, s2,
...])
Hashes fields using basis as a universal hash parameter,
then applies the bundle link selection algorithm to
choose one of the listed slaves represented as
slave_type. Currently the only supported slave_type is
ofport. Thus, each s1 through sN should be an OpenFlow
port number. Outputs to the selected slave.
Currently, fields must be either eth_src, symmetric_l4,
symmetric_l3l4, or symmetric_l3l4+udp, and algorithm must
be one of hrw and active_backup.
Example: bundle(eth_src,0,hrw,ofport,slaves:4,8) uses an
Ethernet source hash with basis 0, to select between
OpenFlow ports 4 and 8 using the Highest Random Weight
algorithm.
Refer to nicira-ext.h for more details.
bundle_load(fields, basis, algorithm, slave_type,
dst[start..end], slaves:[s1, s2, ...])
Has the same behavior as the bundle action, with one
exception. Instead of outputting to the selected slave,
it writes its selection to dst[start..end], which must be
an NXM field as described above.
Example: bundle_load(eth_src, 0, hrw, ofport,
NXM_NX_REG0[], slaves:4, 8) uses an Ethernet source hash
with basis 0, to select between OpenFlow ports 4 and 8
using the Highest Random Weight algorithm, and writes the
selection to NXM_NX_REG0[].
Refer to nicira-ext.h for more details.
learn(argument[,argument]...)
This action adds or modifies a flow in an OpenFlow table,
similar to ovs-ofctl --strict mod-flows. The arguments
specify the flow's match fields, actions, and other prop‐
erties, as follows. At least one match criterion and one
action argument should ordinarily be specified.
idle_timeout=seconds
hard_timeout=seconds
priority=value
cookie=value
send_flow_rem
These arguments have the same meaning as in the
usual ovs-ofctl flow syntax.
fin_idle_timeout=seconds
fin_hard_timeout=seconds
Adds a fin_timeout action with the specified argu‐
ments to the new flow. This feature was added in
Open vSwitch 1.5.90.
table=number
The table in which the new flow should be
inserted. Specify a decimal number between 0 and
254. The default, if table is unspecified, is ta‐
ble 1.
delete_learned
This flag enables deletion of the learned flows
when the flow with the learn action is removed.
Specifically, when the last learn action with this
flag and particular table and cookie values is
removed, the switch deletes all of the flows in
the specified table with the specified cookie.
This flag was added in Open vSwitch 2.4.
field=value
field[start..end]=src[start..end]
field[start..end]
Adds a match criterion to the new flow.
The first form specifies that field must match the
literal value, e.g. dl_type=0x0800. All of the
fields and values for ovs-ofctl flow syntax are
available with their usual meanings.
The second form specifies that field[start..end]
in the new flow must match src[start..end] taken
from the flow currently being processed.
The third form is a shorthand for the second form.
It specifies that field[start..end] in the new
flow must match field[start..end] taken from the
flow currently being processed.
load:value->gt;>gt;dst[start..end]
load:src[start..end]->gt;>gt;dst[start..end]
Adds a load action to the new flow.
The first form loads the literal value into bits
start through end, inclusive, in field dst. Its
syntax is the same as the load action described
earlier in this section.
The second form loads src[start..end], a value
from the flow currently being processed, into bits
start through end, inclusive, in field dst.
output:field[start..end]
Add an output action to the new flow's actions,
that outputs to the OpenFlow port taken from
field[start..end], which must be an NXM field as
described above.
For best performance, segregate learned flows into a ta‐
ble (using table=number) that is not used for any other
flows except possibly for a lowest-priority ``catch-all''
flow, that is, a flow with no match criteria. (This is
why the default table is 1, to keep the learned flows
separate from the primary flow table 0.)
clear_actions
Clears all the actions in the action set immediately.
write_actions([action][,action...])
Add the specific actions to the action set. The syntax
of actions is the same as in the actions= field. The
action set is carried between flow tables and then exe‐
cuted at the end of the pipeline.
The actions in the action set are applied in the follow‐
ing order, as required by the OpenFlow specification,
regardless of the order in which they were added to the
action set. Except as specified otherwise below, the
action set only holds at most a single action of each
type. When more than one action of a single type is
written to the action set, the one written later replaces
the earlier action:
1. strip_vlan
pop_mpls
2. push_mpls
3. push_vlan
4. dec_ttl
dec_mpls_ttl
5. load
move
mod_dl_dst
mod_dl_src
mod_nw_dst
mod_nw_src
mod_nw_tos
mod_nw_ecn
mod_nw_ttl
mod_tp_dst
mod_tp_src
mod_vlan_pcp
mod_vlan_vid
set_field
set_tunnel
set_tunnel64
The action set can contain any number of these
actions, with cumulative effect. They will be
applied in the order as added. That is, when mul‐
tiple actions modify the same part of a field, the
later modification takes effect, and when they
modify different parts of a field (or different
fields), then both modifications are applied.
6. set_queue
7. group
output
resubmit
If more than one of these actions is present, then
the one listed earliest above is executed and the
others are ignored, regardless of the order in
which they were added to the action set. (If none
of these actions is present, the action set has no
real effect, because the modified packet is not
sent anywhere and thus the modifications are not
visible.)
Only the actions listed above may be written to the
action set.
write_metadata:value[/mask]
Updates the metadata field for the flow. If mask is omit‐
ted, the metadata field is set exactly to value; if mask
is specified, then a 1-bit in mask indicates that the
corresponding bit in the metadata field will be replaced
with the corresponding bit from value. Both value and
mask are 64-bit values that are decimal by default; use a
0x prefix to specify them in hexadecimal.
meter:meter_id
Apply the meter_id before any other actions. If a meter
band rate is exceeded, the packet may be dropped, or mod‐
ified, depending on the meter band type. See the descrip‐
tion of the Meter Table Commands, above, for more
details.
goto_table:table
Indicates the next table in the process pipeline.
fin_timeout(argument[,argument])
This action changes the idle timeout or hard timeout, or
both, of this OpenFlow rule when the rule matches a TCP
packet with the FIN or RST flag. When such a packet is
observed, the action reduces the rule's timeouts to those
specified on the action. If the rule's existing timeout
is already shorter than the one that the action speci‐
fies, then that timeout is unaffected.
argument takes the following forms:
idle_timeout=seconds
Causes the flow to expire after the given number
of seconds of inactivity.
hard_timeout=seconds
Causes the flow to expire after the given number
of seconds, regardless of activity. (seconds
specifies time since the flow's creation, not
since the receipt of the FIN or RST.)
This action was added in Open vSwitch 1.5.90.
sample(argument[,argument]...)
Samples packets and sends one sample for every sampled
packet.
argument takes the following forms:
probability=packets
The number of sampled packets out of 65535. Must
be greater or equal to 1.
collector_set_id=id
The unsigned 32-bit integer identifier of the set
of sample collectors to send sampled packets to.
Defaults to 0.
obs_domain_id=id
When sending samples to IPFIX collectors, the
unsigned 32-bit integer Observation Domain ID sent
in every IPFIX flow record. Defaults to 0.
obs_point_id=id
When sending samples to IPFIX collectors, the
unsigned 32-bit integer Observation Point ID sent
in every IPFIX flow record. Defaults to 0.
Refer to ovs-vswitchd.conf.db(8) for more details on con‐
figuring sample collector sets.
This action was added in Open vSwitch 1.10.90.
exit This action causes Open vSwitch to immediately halt exe‐
cution of further actions. Those actions which have
already been executed are unaffected. Any further
actions, including those which may be in other tables, or
different levels of the resubmit call stack, are ignored.
Actions in the action set is still executed (specify
clear_actions before exit to discard them).
conjunction(id, k/n)
An individual OpenFlow flow can match only a single value
for each field. However, situations often arise where
one wants to match one of a set of values within a field
or fields. For matching a single field against a set, it
is straightforward and efficient to add multiple flows to
the flow table, one for each value in the set. For exam‐
ple, one might use the following flows to send packets
with IP source address a, b, c, or d to the OpenFlow con‐
troller:
ip,ip_src=a actions=controller
ip,ip_src=b actions=controller
ip,ip_src=c actions=controller
ip,ip_src=d actions=controller
Similarly, these flows send packets with IP destination
address e, f, g, or h to the OpenFlow controller:
ip,ip_dst=e actions=controller
ip,ip_dst=f actions=controller
ip,ip_dst=g actions=controller
ip,ip_dst=h actions=controller
Installing all of the above flows in a single flow table
yields a disjunctive effect: a packet is sent to the con‐
troller if ip_src ∈ {a,b,c,d} or ip_dst ∈ {e,f,g,h} (or
both). (Pedantically, if both of the above sets of flows
are present in the flow table, they should have different
priorities, because OpenFlow says that the results are
undefined when two flows with same priority can both
match a single packet.)
Suppose, on the other hand, one wishes to match conjunc‐
tively, that is, to send a packet to the controller only
if both ip_src ∈ {a,b,c,d} and ip_dst ∈ {e,f,g,h}. This
requires 4 × 4 = 16 flows, one for each possible pairing
of ip_src and ip_dst. That is acceptable for our small
example, but it does not gracefully extend to larger sets
or greater numbers of dimensions.
The conjunction action is a solution for conjunctive
matches that is built into Open vSwitch. A conjunction
action ties groups of individual OpenFlow flows into
higher-level ``conjunctive flows''. Each group corre‐
sponds to one dimension, and each flow within the group
matches one possible value for the dimension. A packet
that matches one flow from each group matches the con‐
junctive flow.
To implement a conjunctive flow with conjunction, assign
the conjunctive flow a 32-bit id, which must be unique
within an OpenFlow table. Assign each of the n ≥ 2
dimensions a unique number from 1 to n; the ordering is
unimportant. Add one flow to the OpenFlow flow table for
each possible value of each dimension with conjunc‐
tion(id, k/n) as the flow's actions, where k is the num‐
ber assigned to the flow's dimension. Together, these
flows specify the conjunctive flow's match condition.
When the conjunctive match condition is met, Open vSwitch
looks up one more flow that specifies the conjunctive
flow's actions and receives its statistics. This flow is
found by setting conj_id to the specified id and then
again searching the flow table.
The following flows provide an example. Whenever the IP
source is one of the values in the flows that match on
the IP source (dimension 1 of 2), and the IP destination
is one of the values in the flows that match on IP desti‐
nation (dimension 2 of 2), Open vSwitch searches for a
flow that matches conj_id against the conjunction ID
(1234), finding the first flow listed below.
conj_id=1234 actions=controller
ip,ip_src=10.0.0.1 actions=conjunction(1234, 1/2)
ip,ip_src=10.0.0.4 actions=conjunction(1234, 1/2)
ip,ip_src=10.0.0.6 actions=conjunction(1234, 1/2)
ip,ip_src=10.0.0.7 actions=conjunction(1234, 1/2)
ip,ip_dst=10.0.0.2 actions=conjunction(1234, 2/2)
ip,ip_dst=10.0.0.5 actions=conjunction(1234, 2/2)
ip,ip_dst=10.0.0.7 actions=conjunction(1234, 2/2)
ip,ip_dst=10.0.0.8 actions=conjunction(1234, 2/2)
Many subtleties exist:
· In the example above, every flow in a single
dimension has the same form, that is, dimension 1
matches on ip_src, dimension 2 on ip_dst, but this
is not a requirement. Different flows within a
dimension may match on different bits within a
field (e.g. IP network prefixes of different
lengths, or TCP/UDP port ranges as bitwise
matches), or even on entirely different fields
(e.g. to match packets for TCP source port 80 or
TCP destination port 80).
· The flows within a dimension can vary their
matches across more than one field, e.g. to match
only specific pairs of IP source and destination
addresses or L4 port numbers.
· A flow may have multiple conjunction actions, with
different id values. This is useful for multiple
conjunctive flows with overlapping sets. If one
conjunctive flow matches packets with both ip_src
∈ {a,b} and ip_dst ∈ {d,e} and a second conjunc‐
tive flow matches ip_src ∈ {b,c} and ip_dst ∈
{f,g}, for example, then the flow that matches
ip_src=b would have two conjunction actions, one
for each conjunctive flow. The order of conjunc‐
tion actions within a list of actions is not sig‐
nificant.
· A flow with conjunction actions may also include
note actions for annotations, but not any other
kind of actions. (They would not be useful
because they would never be executed.)
· All of the flows that constitute a conjunctive
flow with a given id must have the same priority.
(Flows with the same id but different priorities
are currently treated as different conjunctive
flows, that is, currently id values need only be
unique within an OpenFlow table at a given prior‐
ity. This behavior isn't guaranteed to stay the
same in later releases, so please use id values
unique within an OpenFlow table.)
· Conjunctive flows must not overlap with each
other, at a given priority, that is, any given
packet must be able to match at most one conjunc‐
tive flow at a given priority. Overlapping con‐
junctive flows yield unpredictable results.
· Following a conjunctive flow match, the search for
the flow with conj_id=id is done in the same gen‐
eral-purpose way as other flow table searches, so
one can use flows with conj_id=id to act differ‐
ently depending on circumstances. (One exception
is that the search for the conj_id=id flow itself
ignores conjunctive flows, to avoid recursion.) If
the search with conj_id=id fails, Open vSwitch
acts as if the conjunctive flow had not matched at
all, and continues searching the flow table for
other matching flows.
· OpenFlow prerequisite checking occurs for the flow
with conj_id=id in the same way as any other flow,
e.g. in an OpenFlow 1.1+ context, putting a
mod_nw_src action into the example above would
require adding an ip match, like this:
conj_id=1234,ip actions=mod_nw_src:1.2.3.4,con‐
troller
· OpenFlow prerequisite checking also occurs for the
individual flows that comprise a conjunctive match
in the same way as any other flow.
· The flows that constitute a conjunctive flow do
not have useful statistics. They are never
updated with byte or packet counts, and so on.
(For such a flow, therefore, the idle and hard
timeouts work much the same way.)
· Conjunctive flows can be a useful building block
for negation, that is, inequality matches like
tcp_src ≠ 80. To implement an inequality match,
convert it to a pair of range matches, e.g. 0 ≤
tcp_src 80 and 80 tcp_src ≤ 65535, then con‐
vert each of the range matches into a collection
of bitwise matches as explained above in the
description of tcp_src.
· Sometimes there is a choice of which flows include
a particular match. For example, suppose that we
added an extra constraint to our example, to match
on ip_src ∈ {a,b,c,d} and ip_dst ∈ {e,f,g,h} and
tcp_dst = i. One way to implement this is to add
the new constraint to the conj_id flow, like this:
conj_id=1234,tcp,tcp_dst=i
actions=mod_nw_src:1.2.3.4,controller
but this is not recommended because of the cost of
the extra flow table lookup. Instead, add the
constraint to the individual flows, either in one
of the dimensions or (slightly better) all of
them.
· A conjunctive match must have n ≥ 2 dimensions
(otherwise a conjunctive match is not necessary).
Open vSwitch enforces this.
· Each dimension within a conjunctive match should
ordinarily have more than one flow. Open vSwitch
does not enforce this.
The conjunction action and conj_id field were introduced
in Open vSwitch 2.4.
An opaque identifier called a cookie can be used as a handle to iden‐
tify a set of flows:
cookie=value
A cookie can be associated with a flow using the add-flow,
add-flows, and mod-flows commands. value can be any 64-bit num‐
ber and need not be unique among flows. If this field is omit‐
ted, a default cookie value of 0 is used.
cookie=value/mask
When using NXM, the cookie can be used as a handle for querying,
modifying, and deleting flows. value and mask may be supplied
for the del-flows, mod-flows, dump-flows, and dump-aggregate
commands to limit matching cookies. A 1-bit in mask indicates
that the corresponding bit in cookie must match exactly, and a
0-bit wildcards that bit. A mask of -1 may be used to exactly
match a cookie.
The mod-flows command can update the cookies of flows that match
a cookie by specifying the cookie field twice (once with a mask
for matching and once without to indicate the new value):
ovs-ofctl mod-flows br0 cookie=1,actions=normal
Change all flows' cookies to 1 and change their actions
to normal.
ovs-ofctl mod-flows br0 cookie=1/-1,cookie=2,actions=normal
Update cookies with a value of 1 to 2 and change their
actions to normal.
The ability to match on cookies was added in Open vSwitch 1.5.0.
The following additional field sets the priority for flows added by the
add-flow and add-flows commands. For mod-flows and del-flows when
--strict is specified, priority must match along with the rest of the
flow specification. For mod-flows without --strict, priority is only
significant if the command creates a new flow, that is, non-strict
mod-flows does not match on priority and will not change the priority
of existing flows. Other commands do not allow priority to be speci‐
fied.
priority=value
The priority at which a wildcarded entry will match in compari‐
son to others. value is a number between 0 and 65535, inclu‐
sive. A higher value will match before a lower one. An exact-
match entry will always have priority over an entry containing
wildcards, so it has an implicit priority value of 65535. When
adding a flow, if the field is not specified, the flow's prior‐
ity will default to 32768.
OpenFlow leaves behavior undefined when two or more flows with
the same priority can match a single packet. Some users expect
``sensible'' behavior, such as more specific flows taking prece‐
dence over less specific flows, but OpenFlow does not specify
this and Open vSwitch does not implement it. Users should
therefore take care to use priorities to ensure the behavior
that they expect.
The add-flow, add-flows, and mod-flows commands support the following
additional options. These options affect only new flows. Thus, for
add-flow and add-flows, these options are always significant, but for
mod-flows they are significant only if the command creates a new flow,
that is, their values do not update or affect existing flows.
idle_timeout=seconds
Causes the flow to expire after the given number of seconds of
inactivity. A value of 0 (the default) prevents a flow from
expiring due to inactivity.
hard_timeout=seconds
Causes the flow to expire after the given number of seconds,
regardless of activity. A value of 0 (the default) gives the
flow no hard expiration deadline.
importance=value
Sets the importance of a flow. The flow entry eviction mecha‐
nism can use importance as a factor in deciding which flow to
evict. A value of 0 (the default) makes the flow non-evictable
on the basis of importance. Specify a value between 0 and
65535.
Only OpenFlow 1.4 and later support importance.
send_flow_rem
Marks the flow with a flag that causes the switch to generate a
``flow removed'' message and send it to interested controllers
when the flow later expires or is removed.
check_overlap
Forces the switch to check that the flow match does not overlap
that of any different flow with the same priority in the same
table. (This check is expensive so it is best to avoid it.)
The dump-flows, dump-aggregate, del-flow and del-flows commands support
one additional optional field:
out_port=port
If set, a matching flow must include an output action to port,
which must be an OpenFlow port number or name (e.g. local).
Table Entry Output
The dump-tables and dump-aggregate commands print information about the
entries in a datapath's tables. Each line of output is a flow entry as
described in Flow Syntax, above, plus some additional fields:
duration=secs
The time, in seconds, that the entry has been in the table.
secs includes as much precision as the switch provides, possibly
to nanosecond resolution.
n_packets
The number of packets that have matched the entry.
n_bytes
The total number of bytes from packets that have matched the
entry.
The following additional fields are included only if the switch is Open
vSwitch 1.6 or later and the NXM flow format is used to dump the flow
(see the description of the --flow-format option below). The values of
these additional fields are approximations only and in particular
idle_age will sometimes become nonzero even for busy flows.
hard_age=secs
The integer number of seconds since the flow was added or modi‐
fied. hard_age is displayed only if it differs from the integer
part of duration. (This is separate from duration because
mod-flows restarts the hard_timeout timer without zeroing dura‐
tion.)
idle_age=secs
The integer number of seconds that have passed without any pack‐
ets passing through the flow.
Group Syntax
Some ovs-ofctl commands accept an argument that describes a group or
groups. Such flow descriptions comprise a series field=value assign‐
ments, separated by commas or white space. (Embedding spaces into a
group description normally requires quoting to prevent the shell from
breaking the description into multiple arguments.). Unless noted other‐
wise only the last instance of each field is honoured.
group_id=id
The integer group id of group. When this field is specified in
del-groups or dump-groups, the keyword "all" may be used to des‐
ignate all groups. This field is required.
type=type
The type of the group. The add-group, add-groups and mod-groups
commands require this field. It is prohibited for other com‐
mands. The following keywords designated the allowed types:
all Execute all buckets in the group.
select Execute one bucket in the group. The switch should
select the bucket in such a way that should implement
equal load sharing is achieved. The switch may option‐
ally select the bucket based on bucket weights.
indirect
Executes the one bucket in the group.
ff
fast_failover
Executes the first live bucket in the group which is
associated with a live port or group.
command_bucket_id=id
The bucket to operate on. The insert-buckets and remove-buckets
commands require this field. It is prohibited for other com‐
mands. id may be an integer or one of the following keywords:
all Operate on all buckets in the group. Only valid when
used with the remove-buckets command in which case the
effect is to remove all buckets from the group.
first Operate on the first bucket present in the group. In the
case of the insert-buckets command the effect is to
insert new bucets just before the first bucket already
present in the group; or to replace the buckets of the
group if there are no buckets already present in the
group. In the case of the remove-buckets command the
effect is to remove the first bucket of the group; or do
nothing if there are no buckets present in the group.
last Operate on the last bucket present in the group. In the
case of the insert-buckets command the effect is to
insert new bucets just after the last bucket already
present in the group; or to replace the buckets of the
group if there are no buckets already present in the
group. In the case of the remove-buckets command the
effect is to remove the last bucket of the group; or do
nothing if there are no buckets present in the group.
If id is an integer then it should correspond to the bucket_id
of a bucket present in the group. In case of the insert-buckets
command the effect is to insert buckets just before the bucket
in the group whose bucket_id is id. In case of the iremove-
buckets command the effect is to remove the in the group whose
bucket_id is id. It is an error if there is no bucket persent
group in whose bucket_id is id.
selection_method=method
The selection method used to select a bucket for a select group.
This is a string of 1 to 15 bytes in length known to lower lay‐
ers. This field is optional for add-group, add-groups and
mod-group commands on groups of type select. Prohibited other‐
wise. The default value is the empty string.
This option will use a Netronome OpenFlow extension which is
only supported when using Open vSwitch 2.4 and later with Open‐
Flow 1.5 and later.
selection_method_param=param
64-bit integer parameter to the selection method selected by the
selection_method field. The parameter's use is defined by the
lower-layer that implements the selection_method. It is
optional if the selection_method field is specified as a non-
empty string. Prohibited otherwise. The default value is zero.
This option will use a Netronome OpenFlow extension which is
only supported when using Open vSwitch 2.4 and later with Open‐
Flow 1.5 and later.
fields=param
The field parameters to selection method selected by the selec‐
tion_method field. The syntax is described in Flow Syntax with
the additional restrictions that if a value is provided it is
treated as a wildcard mask and wildcard masks following a slash
are prohibited. The pre-requisites of fields must be provided by
any flows that output to the group. The use of the fields is
defined by the lower-layer that implements the selection_method.
They are optional if the selection_method field is specified as
a non-empty string. Prohibited otherwise. The default is no
fields.
This option will use a Netronome OpenFlow extension which is
only supported when using Open vSwitch 2.4 and later with Open‐
Flow 1.5 and later.
bucket=bucket_parameters
The add-group, add-groups and mod-group commands require at
least one bucket field. Bucket fields must appear after all
other fields. Multiple bucket fields to specify multiple buck‐
ets. The order in which buckets are specified corresponds to
their order in the group. If the type of the group is "indirect"
then only one group may be specified. bucket_parameters con‐
sists of a list of field=value assignments, separated by commas
or white space followed by a comma-separated list of actions.
The fields for bucket_parameters are:
bucket_id=id
The 32-bit integer group id of the bucket. Values
greater than 0xffffff00 are reserved. This field was
added in Open vSwitch 2.4 to conform with the OpenFlow
1.5 specification. It is not supported when earlier ver‐
sions of OpenFlow are used. Open vSwitch will automati‐
cally allocate bucket ids when they are not specified.
actions=[action][,action...]
The syntax of actions are identical to the actions= field
described in Flow Syntax above. Specyfing actions= is
optional, any unknown bucket parameter will be inter‐
preted as an action.
weight=value
The relative weight of the bucket as an integer. This may
be used by the switch during bucket select for groups
whose type is select.
watch_port=port
Port used to determine liveness of group. This or the
watch_group field is required for groups whose type is ff
or fast_failover.
watch_group=group_id
Group identifier of group used to determine liveness of
group. This or the watch_port field is required for
groups whose type is ff or fast_failover.
Meter Syntax
The meter table commands accept an argument that describes a meter.
Such meter descriptions comprise a series field=value assignments, sep‐
arated by commas or white space. (Embedding spaces into a group
description normally requires quoting to prevent the shell from break‐
ing the description into multiple arguments.). Unless noted otherwise
only the last instance of each field is honoured.
meter=id
The integer meter id of the meter. When this field is specified
in del-meter, dump-meter, or meter-stats, the keyword "all" may
be used to designate all meters. This field is required, exept
for meter-stats, which dumps all stats when this field is not
specified.
kbps
pktps The unit for the meter band rate parameters, either kilobits per
second, or packets per second, respectively. One of these must
be specified. The burst size unit corresponds to the rate unit
by dropping the "per second", i.e., burst is in units of kilo‐
bits or packets, respectively.
burst Specify burst size for all bands, or none of them, if this flag
is not given.
stats Collect meter and band statistics.
bands=band_parameters
The add-meter and mod-meter commands require at least one band
specification. Bands must appear after all other fields.
type=type
The type of the meter band. This keyword starts a new
band specification. Each band specifies a rate above
which the band is to take some action. The action depends
on the band type. If multiple bands' rate is exceeded,
then the band with the highest rate among the exceeded
bands is selected. The following keywords designate the
allowed meter band types:
drop Drop packets exceeding the band's rate limit.
The other band_parameters are:
rate=value
The relative rate limit for this band, in kilobits per
second or packets per second, depending on the meter
flags defined above.
burst_size=size
The maximum burst allowed for the band. If pktps is
specified, then size is a packet count, otherwise it is
in kilobits. If unspecified, the switch is free to
select some reasonable value depending on its configura‐
tion.
OPTIONS
--strict
Uses strict matching when running flow modification commands.
--bundle
Execute flow mods as an OpenFlow 1.4 atomic bundle transaction.
· Within a bundle, all flow mods are processed in the order
they appear and as a single atomic transaction, meaning
that if one of them fails, the whole transaction fails
and none of the changes are made to the switch's flow ta‐
ble, and that each given datapath packet traversing the
OpenFlow tables sees the flow tables either as before the
transaction, or after all the flow mods in the bundle
have been successfully applied.
· The beginning and the end of the flow table modification
commands in a bundle are delimited with OpenFlow 1.4 bun‐
dle control messages, which makes it possible to stream
the included commands without explicit OpenFlow barriers,
which are otherwise used after each flow table modifica‐
tion command. This may make large modifications execute
faster as a bundle.
· Bundles require OpenFlow 1.4 or higher. An explicit -O
OpenFlow14 option is not needed, but you may need to
enable OpenFlow 1.4 support for OVS by setting the OVSDB
protocols column in the bridge table.
-O [version[,version]...]
--protocols=[version[,version]...]
Sets the OpenFlow protocol versions that are allowed when estab‐
lishing an OpenFlow session.
The following versions are considered to be ready for general
use. These protocol versions are enabled by default:
· OpenFlow10, for OpenFlow 1.0.
Support for the following protocol versions is provided for
testing and development purposes. They are not enabled by
default:
· OpenFlow11, for OpenFlow 1.1.
· OpenFlow12, for OpenFlow 1.2.
· OpenFlow13, for OpenFlow 1.3.
-F format[,format...]
--flow-format=format[,format...]
ovs-ofctl supports the following individual flow formats, any
number of which may be listed as format:
OpenFlow10-table_id
This is the standard OpenFlow 1.0 flow format. All Open‐
Flow switches and all versions of Open vSwitch support
this flow format.
OpenFlow10+table_id
This is the standard OpenFlow 1.0 flow format plus a
Nicira extension that allows ovs-ofctl to specify the
flow table in which a particular flow should be placed.
Open vSwitch 1.2 and later supports this flow format.
NXM-table_id (Nicira Extended Match)
This Nicira extension to OpenFlow is flexible and exten‐
sible. It supports all of the Nicira flow extensions,
such as tun_id and registers. Open vSwitch 1.1 and later
supports this flow format.
NXM+table_id (Nicira Extended Match)
This combines Nicira Extended match with the ability to
place a flow in a specific table. Open vSwitch 1.2 and
later supports this flow format.
OXM-OpenFlow12
OXM-OpenFlow13
OXM-OpenFlow14
These are the standard OXM (OpenFlow Extensible Match)
flow format in OpenFlow 1.2, 1.3, and 1.4, respectively.
ovs-ofctl also supports the following abbreviations for collec‐
tions of flow formats:
any Any supported flow format.
OpenFlow10
OpenFlow10-table_id or OpenFlow10+table_id.
NXM NXM-table_id or NXM+table_id.
OXM OXM-OpenFlow12, OXM-OpenFlow13, or OXM-OpenFlow14.
For commands that modify the flow table, ovs-ofctl by default
negotiates the most widely supported flow format that supports
the flows being added. For commands that query the flow table,
ovs-ofctl by default uses the most advanced format supported by
the switch.
This option, where format is a comma-separated list of one or
more of the formats listed above, limits ovs-ofctl's choice of
flow format. If a command cannot work as requested using one of
the specified flow formats, ovs-ofctl will report a fatal error.
-P format
--packet-in-format=format
ovs-ofctl supports the following packet_in formats, in order of
increasing capability:
openflow10
This is the standard OpenFlow 1.0 packet in format. It
should be supported by all OpenFlow switches.
nxm (Nicira Extended Match)
This packet_in format includes flow metadata encoded
using the NXM format.
Usually, ovs-ofctl prefers the nxm packet_in format, but will
allow the switch to choose its default if nxm is unsupported.
When format is one of the formats listed in the above table,
ovs-ofctl will insist on the selected format. If the switch
does not support the requested format, ovs-ofctl will report a
fatal error. This option only affects the monitor command.
--timestamp
Print a timestamp before each received packet. This option only
affects the monitor, snoop, and ofp-parse-pcap commands.
-m
--more Increases the verbosity of OpenFlow messages printed and logged
by ovs-ofctl commands. Specify this option more than once to
increase verbosity further.
--sort[=field]
--rsort[=field]
Display output sorted by flow field in ascending (--sort) or
descending (--rsort) order, where field is any of the fields
that are allowed for matching or priority to sort by priority.
When field is omitted, the output is sorted by priority. Spec‐
ify these options multiple times to sort by multiple fields.
Any given flow will not necessarily specify a value for a given
field. This requires special treatement:
· A flow that does not specify any part of a field that is
used for sorting is sorted after all the flows that do
specify the field. For example, --sort=tcp_src will sort
all the flows that specify a TCP source port in ascending
order, followed by the flows that do not specify a TCP
source port at all.
· A flow that only specifies some bits in a field is sorted
as if the wildcarded bits were zero. For example,
--sort=nw_src would sort a flow that specifies
nw_src=192.168.0.0/24 the same as nw_src=192.168.0.0.
These options currently affect only dump-flows output. The fol‐
lowing options are valid on POSIX based platforms.
--pidfile[=pidfile]
Causes a file (by default, ovs-ofctl.pid) to be created indicat‐
ing the PID of the running process. If the pidfile argument is
not specified, or if it does not begin with /, then it is cre‐
ated in /var/run/openvswitch.
If --pidfile is not specified, no pidfile is created.
--overwrite-pidfile
By default, when --pidfile is specified and the specified pid‐
file already exists and is locked by a running process,
ovs-ofctl refuses to start. Specify --overwrite-pidfile to
cause it to instead overwrite the pidfile.
When --pidfile is not specified, this option has no effect.
--detach
Causes ovs-ofctl to detach itself from the foreground session
and run as a background process. ovs-ofctl detaches only when
executing the monitor or snoop commands.
--monitor
Creates an additional process to monitor the ovs-ofctl daemon.
If the daemon dies due to a signal that indicates a programming
error (SIGABRT, SIGALRM, SIGBUS, SIGFPE, SIGILL, SIGPIPE,
SIGSEGV, SIGXCPU, or SIGXFSZ) then the monitor process starts a
new copy of it. If the daemon dies or exits for another reason,
the monitor process exits.
This option is normally used with --detach, but it also func‐
tions without it.
--no-chdir
By default, when --detach is specified, ovs-ofctl changes its
current working directory to the root directory after it
detaches. Otherwise, invoking ovs-ofctl from a carelessly cho‐
sen directory would prevent the administrator from unmounting
the file system that holds that directory.
Specifying --no-chdir suppresses this behavior, preventing
ovs-ofctl from changing its current working directory. This may
be useful for collecting core files, since it is common behavior
to write core dumps into the current working directory and the
root directory is not a good directory to use.
This option has no effect when --detach is not specified.
--user Causes ovs-ofctl to run as a non root user specified in
"user:group", thus dropping all root privileges. Short forms
"user" and ":group" are also allowed, with current user or group
are assumed respectively. Only daemons started by the root user
accepts this argument.
On Linux, daemons will be granted CAP_IPC_LOCK and
CAP_NET_BIND_SERVICES before dropping root privileges. Daemons
interact with datapath, such as ovs-vswitchd, will be granted
two additional capabilities, namely CAP_NET_ADMIN and
CAP_NET_RAW.
On Windows, this option is not currently supported. For security
reasons, specifying this option will cause the daemon process
not to start.
--unixctl=socket
Sets the name of the control socket on which ovs-ofctl listens
for runtime management commands (see RUNTIME MANAGEMENT COM‐
MANDS, below). If socket does not begin with /, it is inter‐
preted as relative to /var/run/openvswitch. If --unixctl is not
used at all, the default socket is /var/run/open‐
vswitch/ovs-ofctl.pid.ctl, where pid is ovs-ofctl's process ID.
On Windows, uses a kernel chosen TCP port on the localhost to
listen for runtime management commands. The kernel chosen TCP
port value is written in a file whose absolute path is pointed
by socket. If --unixctl is not used at all, the file is created
as ovs-ofctl.ctl in the configured OVS_RUNDIR directory.
Specifying none for socket disables the control socket feature.
Public Key Infrastructure Options
-p privkey.pem
--private-key=privkey.pem
Specifies a PEM file containing the private key used as
ovs-ofctl's identity for outgoing SSL connections.
-c cert.pem
--certificate=cert.pem
Specifies a PEM file containing a certificate that certifies the
private key specified on -p or --private-key to be trustworthy.
The certificate must be signed by the certificate authority (CA)
that the peer in SSL connections will use to verify it.
-C cacert.pem
--ca-cert=cacert.pem
Specifies a PEM file containing the CA certificate that
ovs-ofctl should use to verify certificates presented to it by
SSL peers. (This may be the same certificate that SSL peers use
to verify the certificate specified on -c or --certificate, or
it may be a different one, depending on the PKI design in use.)
-C none
--ca-cert=none
Disables verification of certificates presented by SSL peers.
This introduces a security risk, because it means that certifi‐
cates cannot be verified to be those of known trusted hosts.
-v[spec]
--verbose=[spec]
Sets logging levels. Without any spec, sets the log level for
every module and destination to dbg. Otherwise, spec is a list
of words separated by spaces or commas or colons, up to one from
each category below:
· A valid module name, as displayed by the vlog/list com‐
mand on ovs-appctl(8), limits the log level change to the
specified module.
· syslog, console, or file, to limit the log level change
to only to the system log, to the console, or to a file,
respectively.
On Windows platform, syslog is accepted as a word and is
only useful along with the --syslog-target option (the
word has no effect otherwise).
· off, emer, err, warn, info, or dbg, to control the log
level. Messages of the given severity or higher will be
logged, and messages of lower severity will be filtered
out. off filters out all messages. See ovs-appctl(8)
for a definition of each log level.
Case is not significant within spec.
Regardless of the log levels set for file, logging to a file
will not take place unless --log-file is also specified (see
below).
For compatibility with older versions of OVS, any is accepted as
a word but has no effect.
-v
--verbose
Sets the maximum logging verbosity level, equivalent to --ver‐
bose=dbg.
-vPATTERN:destination:pattern
--verbose=PATTERN:destination:pattern
Sets the log pattern for destination to pattern. Refer to
ovs-appctl(8) for a description of the valid syntax for pattern.
-vFACILITY:facility
--verbose=FACILITY:facility
Sets the RFC5424 facility of the log message. facility can be
one of kern, user, mail, daemon, auth, syslog, lpr, news, uucp,
clock, ftp, ntp, audit, alert, clock2, local0, local1, local2,
local3, local4, local5, local6 or local7. If this option is not
specified, daemon is used as the default for the local system
syslog and local0 is used while sending a message to the target
provided via the --syslog-target option.
--log-file[=file]
Enables logging to a file. If file is specified, then it is
used as the exact name for the log file. The default log file
name used if file is omitted is /var/log/open‐
vswitch/ovs-ofctl.log.
--syslog-target=host:port
Send syslog messages to UDP port on host, in addition to the
system syslog. The host must be a numerical IP address, not a
hostname.
--syslog-method=method
Specify method how syslog messages should be sent to syslog dae‐
mon. Following forms are supported:
· libc, use libc syslog() function. This is the default
behavior. Downside of using this options is that libc
adds fixed prefix to every message before it is actually
sent to the syslog daemon over /dev/log UNIX domain
socket.
· unix:file, use UNIX domain socket directly. It is possi‐
ble to specify arbitrary message format with this option.
However, rsyslogd 8.9 and older versions use hard coded
parser function anyway that limits UNIX domain socket
use. If you want to use arbitrary message format with
older rsyslogd versions, then use UDP socket to localhost
IP address instead.
· udp:ip:port, use UDP socket. With this method it is pos‐
sible to use arbitrary message format also with older
rsyslogd. When sending syslog messages over UDP socket
extra precaution needs to be taken into account, for
example, syslog daemon needs to be configured to listen
on the specified UDP port, accidental iptables rules
could be interfering with local syslog traffic and there
are some security considerations that apply to UDP sock‐
ets, but do not apply to UNIX domain sockets.
-h
--help Prints a brief help message to the console.
-V
--version
Prints version information to the console.
RUNTIME MANAGEMENT COMMANDS
ovs-appctl(8) can send commands to a running ovs-ofctl process. The
supported commands are listed below.
exit Causes ovs-ofctl to gracefully terminate. This command applies
only when executing the monitor or snoop commands.
ofctl/set-output-file file
Causes all subsequent output to go to file instead of stderr.
This command applies only when executing the monitor or snoop
commands.
ofctl/send ofmsg...
Sends each ofmsg, specified as a sequence of hex digits that
express an OpenFlow message, on the OpenFlow connection. This
command is useful only when executing the monitor command.
ofctl/barrier
Sends an OpenFlow barrier request on the OpenFlow connection and
waits for a reply. This command is useful only for the monitor
command.
EXAMPLES
The following examples assume that ovs-vswitchd has a bridge named br0
configured.
ovs-ofctl dump-tables br0
Prints out the switch's table stats. (This is more interesting
after some traffic has passed through.)
ovs-ofctl dump-flows br0
Prints the flow entries in the switch.
SEE ALSO
ovs-appctl(8), ovs-vswitchd(8) ovs-vswitchd.conf.db(8)
Open vSwitch 2.4.90 ovs-ofctl(8)