DOKK / manpages / debian 10 / openvswitch-common / ovs-vswitchd.8.en
ovs-vswitchd(8) Open vSwitch Manual ovs-vswitchd(8)

ovs-vswitchd - Open vSwitch daemon

ovs-vswitchd [database]

A daemon that manages and controls any number of Open vSwitch switches on the local machine.

The database argument specifies how ovs-vswitchd connects to ovsdb-server. database may be an OVSDB active or passive connection method, as described in ovsdb(7). The default is unix:/var/run/openvswitch/db.sock.

ovs-vswitchd retrieves its configuration from database at startup. It sets up Open vSwitch datapaths and then operates switching across each bridge described in its configuration files. As the database changes, ovs-vswitchd automatically updates its configuration to match.

ovs-vswitchd switches may be configured with any of the following features:

  • L2 switching with MAC learning.
  • NIC bonding with automatic fail-over and source MAC-based TX load balancing ("SLB").
  • 802.1Q VLAN support.
  • Port mirroring, with optional VLAN tagging.
  • NetFlow v5 flow logging.
  • sFlow(R) monitoring.
  • Connectivity to an external OpenFlow controller, such as NOX.

Only a single instance of ovs-vswitchd is intended to run at a time. A single ovs-vswitchd can manage any number of switch instances, up to the maximum number of supported Open vSwitch datapaths.

ovs-vswitchd does all the necessary management of Open vSwitch datapaths itself. Thus, ovs-dpctl(8) (and its userspace datapath counterparts accessible via ovs-appctl dpctl/command) are not needed with ovs-vswitchd and should not be used because they can interfere with its operation. These tools are still useful for diagnostics.

An Open vSwitch datapath kernel module must be loaded for ovs-vswitchd to be useful. Refer to the documentation for instructions on how to build and load the Open vSwitch kernel module.

Causes ovs-vswitchd to call the mlockall() function, to attempt to lock all of its process memory into physical RAM, preventing the kernel from paging any of its memory to disk. This helps to avoid networking interruptions due to system memory pressure.
Some systems do not support mlockall() at all, and other systems only allow privileged users, such as the superuser, to use it. ovs-vswitchd emits a log message if mlockall() is unavailable or unsuccessful.

For details on initializing the ovs-vswitchd DPDK datapath, refer to the documentation or ovs-vswitchd.conf.db(5) for details.

The following options are valid on POSIX based platforms.

Causes a file (by default, ovs-vswitchd.pid) to be created indicating the PID of the running process. If the pidfile argument is not specified, or if it does not begin with /, then it is created in /var/run/openvswitch.
If --pidfile is not specified, no pidfile is created.
By default, when --pidfile is specified and the specified pidfile already exists and is locked by a running process, ovs-vswitchd refuses to start. Specify --overwrite-pidfile to cause it to instead overwrite the pidfile.
When --pidfile is not specified, this option has no effect.
Runs ovs-vswitchd as a background process. The process forks, and in the child it starts a new session, closes the standard file descriptors (which has the side effect of disabling logging to the console), and changes its current directory to the root (unless --no-chdir is specified). After the child completes its initialization, the parent exits. ovs-vswitchd detaches only after it has connected to the database, retrieved the initial configuration, and set up that configuration.
Creates an additional process to monitor the ovs-vswitchd 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 functions without it.
By default, when --detach is specified, ovs-vswitchd changes its current working directory to the root directory after it detaches. Otherwise, invoking ovs-vswitchd from a carelessly chosen directory would prevent the administrator from unmounting the file system that holds that directory.
Specifying --no-chdir suppresses this behavior, preventing ovs-vswitchd 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.
By default daemon will try to self-confine itself to work with files under well-know, at build-time whitelisted directories. It is better to stick with this default behavior and not to use this flag unless some other Access Control is used to confine daemon. Note that in contrast to other access control implementations that are typically enforced from kernel-space (e.g. DAC or MAC), self-confinement is imposed from the user-space daemon itself and hence should not be considered as a full confinement strategy, but instead should be viewed as an additional layer of security.
Causes ovs-vswitchd to run as a different user specified in "user:group", thus dropping most of the 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 that interact with a datapath, such as ovs-vswitchd, will be granted three additional capabilities, namely CAP_NET_ADMIN, CAP_NET_BROADCAST and CAP_NET_RAW. The capability change will apply even if the new user is root.
On Windows, this option is not currently supported. For security reasons, specifying this option will cause the daemon process not to start.

The following options are valid only on Windows platform.

Causes ovs-vswitchd to run as a service in the background. The service should already have been created through external tools like SC.exe.
Causes the ovs-vswitchd service to be automatically restarted by the Windows services manager if the service dies or exits for unexpected reasons.
When --service is not specified, this option has no effect.


Specifies a PEM file containing the private key used as ovs-vswitchd's identity for outgoing SSL connections.

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.

Specifies a PEM file containing the CA certificate that ovs-vswitchd 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.)

Disables verification of certificates presented by SSL peers. This introduces a security risk, because it means that certificates cannot be verified to be those of known trusted hosts.
When cacert.pem exists, this option has the same effect as -C or --ca-cert. If it does not exist, then ovs-vswitchd will attempt to obtain the CA certificate from the SSL peer on its first SSL connection and save it to the named PEM file. If it is successful, it will immediately drop the connection and reconnect, and from then on all SSL connections must be authenticated by a certificate signed by the CA certificate thus obtained.
This option exposes the SSL connection to a man-in-the-middle attack obtaining the initial CA certificate, but it may be useful for bootstrapping.
This option is only useful if the SSL peer sends its CA certificate as part of the SSL certificate chain. The SSL protocol does not require the server to send the CA certificate.
This option is mutually exclusive with -C and --ca-cert.
Specifies a PEM file that contains one or more additional certificates to send to SSL peers. peer-cacert.pem should be the CA certificate used to sign ovs-vswitchd's own certificate, that is, the certificate specified on -c or --certificate. If ovs-vswitchd's certificate is self-signed, then --certificate and --peer-ca-cert should specify the same file.
This option is not useful in normal operation, because the SSL peer must already have the CA certificate for the peer to have any confidence in ovs-vswitchd's identity. However, this offers a way for a new installation to bootstrap the CA certificate on its first SSL connection.


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 command 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. (If --detach is specified, ovs-vswitchd closes its standard file descriptors, so logging to the console will have no effect.)
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.

Sets the maximum logging verbosity level, equivalent to --verbose=dbg.

Sets the log pattern for destination to pattern. Refer to ovs-appctl(8) for a description of the valid syntax for pattern.

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.
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/openvswitch/ovs-vswitchd.log.
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.
Specify method how syslog messages should be sent to syslog daemon. 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 possible 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 possible 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 sockets, but do not apply to UNIX domain sockets.

Sets the name of the control socket on which ovs-vswitchd listens for runtime management commands (see RUNTIME MANAGEMENT COMMANDS, below). If socket does not begin with /, it is interpreted as relative to /var/run/openvswitch. If --unixctl is not used at all, the default socket is /var/run/openvswitch/ovs-vswitchd.pid.ctl, where pid is ovs-vswitchd's process ID.
On Windows a local named pipe is used to listen for runtime management commands. A file is created in the absolute path as pointed by socket or if --unixctl is not used at all, a file is created as ovs-vswitchd.ctl in the configured OVS_RUNDIR directory. The file exists just to mimic the behavior of a Unix domain socket.
Specifying none for socket disables the control socket feature.

Prints a brief help message to the console.

Prints version information to the console.

ovs-appctl(8) can send commands to a running ovs-vswitchd process. The currently supported commands are described below. The command descriptions assume an understanding of how to configure Open vSwitch.

Causes ovs-vswitchd to gracefully terminate. If --cleanup is specified, release datapath resources configured by ovs-vswitchd. Otherwise, datapath flows and other resources remains undeleted.
Queries the interface for a list of Quality of Service types that are configurable via Open vSwitch for the given interface.
Queries the kernel for Quality of Service configuration and statistics associated with the given interface.
Displays detailed information about Bidirectional Forwarding Detection configured on interface. If interface is not specified, then displays detailed information about all interfaces with BFD enabled.
Force the fault status of the BFD module on interface (or all interfaces if none is given) to be status. status can be "true", "false", or "normal" which reverts to the standard behavior.
Displays detailed information about Connectivity Fault Management configured on interface. If interface is not specified, then displays detailed information about all interfaces with CFM enabled.
Force the fault status of the CFM module on interface (or all interfaces if none is given) to be status. status can be "true", "false", or "normal" which reverts to the standard behavior.
Forces a topology change event on bridge if it's running STP. This may cause it to send Topology Change Notifications to its peers and flush its MAC table. If no bridge is given, forces a topology change event on all bridges.
Displays detailed information about spanning tree on the bridge. If bridge is not specified, then displays detailed information about all bridges with STP enabled.
Forces a topology change event on bridge if it's running RSTP. This may cause it to send Topology Change Notifications to its peers and flush its MAC table. If no bridge is given, forces a topology change event on all bridges.
Displays detailed information about rapid spanning tree on the bridge. If bridge is not specified, then displays detailed information about all bridges with RSTP enabled.

These commands manage bridges.

Flushes bridge MAC address learning table, or all learning tables if no bridge is given.
Lists each MAC address/VLAN pair learned by the specified bridge, along with the port on which it was learned and the age of the entry, in seconds.
Clear bridge MAC address learning table statistics, or all statistics if no bridge is given.
Show MAC address learning table statistics for the specified bridge.
Flushes bridge multicast snooping table, or all snooping tables if no bridge is given.
Lists each multicast group/VLAN pair learned by the specified bridge, along with the port on which it was learned and the age of the entry, in seconds.
Makes bridge drop all of its OpenFlow controller connections and reconnect. If bridge is not specified, then all bridges drop their controller connections and reconnect.
This command might be useful for debugging OpenFlow controller issues.
Lists all flows in bridge, including those normally hidden to commands such as ovs-ofctl dump-flows. Flows set up by mechanisms such as in-band control and fail-open are hidden from the controller since it is not allowed to modify or override them.

These commands manage bonded ports on an Open vSwitch's bridges. To understand some of these commands, it is important to understand a detail of the bonding implementation called ``source load balancing'' (SLB). Instead of directly assigning Ethernet source addresses to slaves, the bonding implementation computes a function that maps an 48-bit Ethernet source addresses into an 8-bit value (a ``MAC hash'' value). All of the Ethernet addresses that map to a single 8-bit value are then assigned to a single slave.

Lists all of the bonds, and their slaves, on each bridge.
Lists all of the bond-specific information (updelay, downdelay, time until the next rebalance) about the given bonded port, or all bonded ports if no port is given. Also lists information about each slave: whether it is enabled or disabled, the time to completion of an updelay or downdelay if one is in progress, whether it is the active slave, the hashes assigned to the slave. Any LACP information related to this bond may be found using the lacp/show command.
Only valid for SLB bonds. Assigns a given MAC hash to a new slave. port specifies the bond port, hash the MAC hash to be migrated (as a decimal number between 0 and 255), and slave the new slave to be assigned.
The reassignment is not permanent: rebalancing or fail-over will cause the MAC hash to be shifted to a new slave in the usual manner.
A MAC hash cannot be migrated to a disabled slave.
Sets slave as the active slave on port. slave must currently be enabled.
The setting is not permanent: a new active slave will be selected if slave becomes disabled.

Enables (or disables) slave on the given bond port, skipping any updelay (or downdelay).
This setting is not permanent: it persists only until the carrier status of slave changes.
Returns the hash value which would be used for mac with vlan and basis if specified.
Lists all of the LACP related information about the given port: active or passive, aggregation key, system id, and system priority. Also lists information about each slave: whether it is enabled or disabled, whether it is attached or detached, port id and priority, actor information, and partner information. If port is not specified, then displays detailed information about all interfaces with CFM enabled.
Lists various stats about LACP PDUs (number of RX/TX PDUs, bad PDUs received) and slave state (number of time slave's state expired/defaulted and carrier status changed) for the given port. If port is not specified, then displays stats of all interfaces with LACP enabled.

The primary way to configure ovs-vswitchd is through the Open vSwitch database, e.g. using ovs-vsctl(8). These commands provide a debugging interface for managing datapaths. They implement the same features (and syntax) as ovs-dpctl(8). Unlike ovs-dpctl(8), these commands work with datapaths that are integrated into ovs-vswitchd (e.g. the netdev datapath type).

Do not use commands to add or remove or modify datapaths if ovs-vswitchd is running because this interferes with ovs-vswitchd's own datapath management.

Creates datapath dp, with a local port also named dp. This will fail if a network device dp already exists.
If netdevs are specified, ovs-vswitchd adds them to the new datapath, just as if add-if was specified.
Deletes datapath dp. If dp is associated with any network devices, they are automatically removed.
Adds each netdev to the set of network devices datapath dp monitors, where dp is the name of an existing datapath, and netdev is the name of one of the host's network devices, e.g. eth0. Once a network device has been added to a datapath, the datapath has complete ownership of the network device's traffic and the network device appears silent to the rest of the system.
A netdev may be followed by a comma-separated list of options. The following options are currently supported:
Specifies the type of port to add. The default type is system.
Requests a specific port number within the datapath. If this option is not specified then one will be automatically assigned.
Adds an arbitrary key-value option to the port's configuration.
ovs-vswitchd.conf.db(5) documents the available port types and options.
Reconfigures each port in dp as specified. An option of the form key=value adds the specified key-value option to the port or overrides an existing key's value. An option of the form key=, that is, without a value, deletes the key-value named key. The type and port number of a port cannot be changed, so type and port_no are only allowed if they match the existing configuration.
Removes each netdev from the list of network devices datapath dp monitors.
Prints the name of each configured datapath on a separate line.
Prints a summary of configured datapaths, including their datapath numbers and a list of ports connected to each datapath. (The local port is identified as port 0.) If -s or --statistics is specified, then packet and byte counters are also printed for each port.
The datapath numbers consists of flow stats and mega flow mask stats.
The "lookups" row displays three stats related to flow lookup triggered by processing incoming packets in the datapath. "hit" displays number of packets matches existing flows. "missed" displays the number of packets not matching any existing flow and require user space processing. "lost" displays number of packets destined for user space process but subsequently dropped before reaching userspace. The sum of "hit" and "miss" equals to the total number of packets datapath processed.
The "flows" row displays the number of flows in datapath.
The "masks" row displays the mega flow mask stats. This row is omitted for datapath not implementing mega flow. "hit" displays the total number of masks visited for matching incoming packets. "total" displays number of masks in the datapath. "hit/pkt" displays the average number of masks visited per packet; the ratio between "hit" and total number of packets processed by the datapath.
If one or more datapaths are specified, information on only those datapaths are displayed. Otherwise, ovs-vswitchd displays information about all configured datapaths.

The following commands are primarily useful for debugging Open vSwitch. The flow table entries (both matches and actions) that they work with are not OpenFlow flow entries. Instead, they are different and considerably simpler flows maintained by the Open vSwitch kernel module. Do not use commands to add or remove or modify datapath flows if ovs-vswitchd is running because it interferes with ovs-vswitchd's own datapath flow management. Use ovs-ofctl(8), instead, to work with OpenFlow flow entries.

The dp argument to each of these commands is optional when exactly one datapath exists, in which case that datapath is the default. When multiple datapaths exist, then a datapath name is required.

Prints to the console all flow entries in datapath dp's flow table. Without -m or --more, output omits match fields that a flow wildcards entirely; with -m or --more, output includes all wildcarded fields.
If filter=filter is specified, only displays the flows that match the filter. filter is a flow in the form similiar to that accepted by ovs-ofctl(8)'s add-flow command. (This is not an OpenFlow flow: besides other differences, it never contains wildcards.) The filter is also useful to match wildcarded fields in the datapath flow. As an example, filter='tcp,tp_src=100' will match the datapath flow containing 'tcp(src=80/0xff00,dst=8080/0xff)'.
If type=type is specified, only displays flows of a specific type. type can be offloaded to display only rules offloaded to the HW or ovs to display only rules from the OVS tables. By default all rules are displayed.
Adds or modifies a flow in dp's flow table that, when a packet matching flow arrives, causes actions to be executed.
The add-flow command succeeds only if flow does not already exist in dp. Contrariwise, mod-flow without --may-create only modifies the actions for an existing flow. With --may-create, mod-flow will add a new flow or modify an existing one.
If -s or --statistics is specified, then mod-flow prints the modified flow's statistics. A flow's statistics are the number of packets and bytes that have passed through the flow, the elapsed time since the flow last processed a packet (if ever), and (for TCP flows) the union of the TCP flags processed through the flow.
With --clear, mod-flow zeros out the flow's statistics. The statistics printed if -s or --statistics is also specified are those from just before clearing the statistics.
NOTE: flow and actions do not match the syntax used with ovs-ofctl(8)'s add-flow command.
Usage Examples

Forward ARP between ports 1 and 2 on datapath myDP:

ovs-dpctl add-flow myDP \
"in_port(1),eth(),eth_type(0x0806),arp()" 2
ovs-dpctl add-flow myDP \
"in_port(2),eth(),eth_type(0x0806),arp()" 1

Forward all IPv4 traffic between two addresses on ports 1 and 2:

ovs-dpctl add-flow myDP \
"in_port(1),eth(),eth_type(0x800),\
ipv4(src=172.31.110.4,dst=172.31.110.5)" 2
ovs-dpctl add-flow myDP \
"in_port(2),eth(),eth_type(0x800),\
ipv4(src=172.31.110.5,dst=172.31.110.4)" 1
Deletes the flow from dp's flow table that matches flow. If -s or --statistics is specified, then del-flow prints the deleted flow's statistics.
Fetches the flow from dp's flow table with unique identifier ufid. ufid must be specified as a string of 32 hexadecimal characters.
Deletes all flow entries from datapath dp's flow table.

The following commands are useful for debugging and configuring the connection tracking table in the datapath.

The dp argument to each of these commands is optional when exactly one datapath exists, in which case that datapath is the default. When multiple datapaths exist, then a datapath name is required.

N.B.(Linux specific): the system datapaths (i.e. the Linux kernel module Open vSwitch datapaths) share a single connection tracking table (which is also used by other kernel subsystems, such as iptables, nftables and the regular host stack). Therefore, the following commands do not apply specifically to one datapath.

Prints to the console all the connection entries in the tracker used by dp. If zone=zone is specified, only shows the connections in zone. With --more, some implementation specific details are included. With --statistics timeouts and timestamps are added to the output.
Flushes the connection entries in the tracker used by dp based on zone and connection tracking tuple ct-tuple. If ct-tuple is not provided, flushes all the connection entries. If zone=zone is specified, only flushes the connections in zone.
If ct-tuple is provided, flushes the connection entry specified by ct-tuple in zone. The zone defaults to 0 if it is not provided. The userspace connection tracker requires flushing with the original pre-NATed tuple and a warning log will be otherwise generated. An example of an IPv4 ICMP ct-tuple:
"ct_nw_src=10.1.1.1,ct_nw_dst=10.1.1.2,ct_nw_proto=1,icmp_type=8,icmp_code=0,icmp_id=10"
An example of an IPv6 TCP ct-tuple:
"ct_ipv6_src=fc00::1,ct_ipv6_dst=fc00::2,ct_nw_proto=6,ct_tp_src=1,ct_tp_dst=2"
Displays the number of connections grouped by protocol used by dp. If zone=zone is specified, numbers refer to the connections in zone. With --more, groups by connection state for each protocol.
For each conntrack bucket, displays the number of connections used by dp. If gt=threshold is specified, bucket numbers are displayed when the number of connections in a bucket is greater than threshold.
Sets the maximum limit of connection tracker entries to maxconns on dp. This can be used to reduce the processing load on the system due to connection tracking or simply limiting connection tracking. If the number of connections is already over the new maximum limit request then the new maximum limit will be enforced when the number of connections decreases to that limit, which normally happens due to connection expiry. Only supported for userspace datapath.
Prints the maximum limit of connection tracker entries on dp. Only supported for userspace datapath.
Prints the current number of connection tracker entries on dp. Only supported for userspace datapath.
Sets the maximum allowed number of connections in a connection tracking zone. A specific zone may be set to limit, and multiple zones may be specified with a comma-separated list. If a per-zone limit for a particular zone is not specified in the datapath, it defaults to the default per-zone limit. A default zone may be specified with the default=default_limit argument. Initially, the default per-zone limit is unlimited. An unlimited number of entries may be set with 0 limit. Only supported for Linux kernel datapath.
Deletes the connection tracking limit for zone. Multiple zones may be specified with a comma-separated list. Only supported for Linux kernel datapath.
Retrieves the maximum allowed number of connections and current counts per-zone. If zone is given, only the specified zone(s) are printed. If no zones are specified, all the zone limits and counts are provided. The command always displays the default zone limit. Only supported for Linux kernel datapath.

These commands are used to expose internal information (mostly statistics) about the "dpif-netdev" userspace datapath. If there is only one datapath (as is often the case, unless dpctl/ commands are used), the dp argument can be omitted. By default the commands present data for all pmd threads in the datapath. By specifying the "-pmd Core" option one can filter the output for a single pmd in the datapath.

Shows performance statistics for one or all pmd threads of the datapath dp. The special thread "main" sums up the statistics of every non pmd thread.

The sum of "emc hits", "smc hits", "megaflow hits" and "miss" is the number of packet lookups performed by the datapath. Beware that a recirculated packet experiences one additional lookup per recirculation, so there may be more lookups than forwarded packets in the datapath.

Cycles are counted using the TSC or similar facilities (when available on the platform). The duration of one cycle depends on the processing platform.

"idle cycles" refers to cycles spent in PMD iterations not forwarding any any packets. "processing cycles" refers to cycles spent in PMD iterations forwarding at least one packet, including the cost for polling, processing and transmitting said packets.

To reset these counters use dpif-netdev/pmd-stats-clear.

Resets to zero the per pmd thread performance numbers shown by the dpif-netdev/pmd-stats-show and dpif-netdev/pmd-perf-show commands. It will NOT reset datapath or bridge statistics, only the values shown by the above commands.
Shows detailed performance metrics for one or all pmds threads of the user space datapath.

The collection of detailed statistics can be controlled by a new configuration parameter "other_config:pmd-perf-metrics". By default it is disabled. The run-time overhead, when enabled, is in the order of 1%.

—
used cycles
—
forwared packets
—
number of rx batches
—
packets/rx batch
—
max. vhostuser queue fill level
—
number of upcalls
—
cycles spent in upcalls
This raw recorded data is used threefold:

1.
In histograms for each of the following metrics:
—
cycles/iteration (logarithmic)
—
packets/iteration (logarithmic)
—
cycles/packet
—
packets/batch
—
max. vhostuser qlen (logarithmic)
—
upcalls
—
cycles/upcall (logarithmic) The histograms bins are divided linear or logarithmic.
2.
A cyclic history of the above metrics for 1024 iterations
3.
A cyclic history of the cummulative/average values per millisecond wall clock for the last 1024 milliseconds:
—
number of iterations
—
avg. cycles/iteration
—
packets (Kpps)
—
avg. packets/batch
—
avg. max vhost qlen
—
upcalls
—
avg. cycles/upcall
The command options are:
Suppress the histograms
Display the last iter_len iteration stats
Display the last ms_len millisecond stats
The output always contains the following global PMD statistics:
Time: 15:24:55.270 .br Measurement duration: 1.008 s

pmd thread numa_id 0 core_id 1:


Cycles: 2419034712 (2.40 GHz)
Iterations: 572817 (1.76 us/it)
- idle: 486808 (15.9 % cycles)
- busy: 86009 (84.1 % cycles)
Rx packets: 2399607 (2381 Kpps, 848 cycles/pkt)
Datapath passes: 3599415 (1.50 passes/pkt)
- EMC hits: 336472 ( 9.3 %)
- SMC hits: 0 ( 0.0 %)
- Megaflow hits: 3262943 (90.7 %, 1.00 subtbl lookups/hit)
- Upcalls: 0 ( 0.0 %, 0.0 us/upcall)
- Lost upcalls: 0 ( 0.0 %)
Tx packets: 2399607 (2381 Kpps)
Tx batches: 171400 (14.00 pkts/batch)

Here "Rx packets" actually reflects the number of packets forwarded by the datapath. "Datapath passes" matches the number of packet lookups as reported by the dpif-netdev/pmd-stats-show command.

To reset the counters and start a new measurement use dpif-netdev/pmd-stats-clear.

The userspace "netdev" datapath is able to supervise the PMD performance metrics and detect iterations with suspicious statistics according to the following criteria:
—
The iteration lasts longer than usec microseconds (default 250). This can be used to capture events where a PMD is blocked or interrupted for such a period of time that there is a risk for dropped packets on any of its Rx queues.
—
The max vhost qlen exceeds a threshold qlen (default 128). This can be used to infer virtio queue overruns and dropped packets inside a VM, which are not visible in OVS otherwise.
Such suspicious iterations can be logged together with their iteration statistics in the ovs-vswitchd.log to be able to correlate them to packet drop or other events outside OVS.

The above command enables (on) or disables (off) supervision and logging at run-time and can be used to adjust the above thresholds for detecting suspicious iterations. By default supervision and logging is disabled.

The command options are:

The number of iterations before the suspicious iteration to be logged (default 5).
The number of iterations after the suspicious iteration to be logged (default 5).
Extend logging interval if another suspicious iteration is detected before logging occurs.
Do not extend logging interval if another suspicious iteration is detected before logging occurs (default).
Suspicious vhost queue fill level threshold. Increase this to 512 if the Qemu supports 1024 virtio queue length (default 128).
Change the duration threshold for a suspicious iteration (default 250 us).

Note: Logging of suspicious iterations itself consumes a considerable amount of processing cycles of a PMD which may be visible in the iteration history. In the worst case this can lead OVS to detect another suspicious iteration caused by logging.

If more than 100 iterations around a suspicious iteration have been logged once, OVS falls back to the safe default values (-b 5 -a 5 -ne) to avoid that logging itself continuously causes logging of further suspicious iterations.

For one or all pmd threads of the datapath dp show the list of queue-ids with port names, which this thread polls.
Reassigns rxqs to pmds in the datapath dp based on their current usage.

These commands manage DPDK related ports (type=dpdk*).

Change the admin state for DPDK interface to up or down. If interface is not specified, then it applies to all DPDK ports.
Detaches device with corresponding pci-address from DPDK. This command can be used to detach device if it wasn't detached automatically after port deletion. Refer to the documentation for details and instructions.
Prints the debug information about memory pool used by DPDK interface. If called without arguments, information of all the available mempools will be printed. For additional mempool statistics enable CONFIG_RTE_LIBRTE_MEMPOOL_DEBUG while building DPDK.

These commands query and modify datapaths. They are are similar to ovs-dpctl(8) commands. dpif/show has the additional functionality, beyond dpctl/show of printing OpenFlow port numbers. The other commands are redundant and will be removed in a future release.

Prints the name of each configured datapath on a separate line.
Prints a summary of configured datapaths, including statistics and a list of connected ports. The port information includes the OpenFlow port number, datapath port number, and the type. (The local port is identified as OpenFlow port 65534.)
Prints to the console all flow entries in datapath dp's flow table. Without -m, output omits match fields that a flow wildcards entirely; with -m output includes all wildcarded fields.
This command is primarily useful for debugging Open vSwitch. The flow table entries that it displays are not OpenFlow flow entries. Instead, they are different and considerably simpler flows maintained by the datapath module. If you wish to see the OpenFlow flow entries, use ovs-ofctl dump-flows.
Deletes all flow entries from datapath dp's flow table and underlying datapath implementation (e.g., kernel datapath module).
This command is primarily useful for debugging Open vSwitch. As discussed in dpif/dump-flows, these entries are not OpenFlow flow entries.

These commands manage the core OpenFlow switch implementation (called ofproto).

Lists the names of the running ofproto instances. These are the names that may be used on ofproto/trace.



Traces the path of an imaginary packet through switch and reports the path that it took. The initial treatment of the packet varies based on the command:
  • ofproto/trace looks the packet up in the OpenFlow flow table, as if the packet had arrived on an OpenFlow port.
  • ofproto/trace-packet-out applies the specified OpenFlow actions, as if the packet, flow, and actions had been specified in an OpenFlow ``packet-out'' request.
The packet's headers (e.g. source and destination) and metadata (e.g. input port), together called its ``flow,'' are usually all that matter for the purpose of tracing a packet. You can specify the flow in the following ways:
odp_flow is a flow in the form printed by ovs-dpctl(8)'s dump-flows command. If all of your bridges have the same type, which is the common case, then you can omit dpname, but if you have bridges of different types (say, both ovs-netdev and ovs-system), then you need to specify a dpname to disambiguate.
br_flow is a flow in the form similar to that accepted by ovs-ofctl(8)'s add-flow command. (This is not an OpenFlow flow: besides other differences, it never contains wildcards.) bridge names of the bridge through which br_flow should be traced.
These commands support the following options:
Generate a packet from the flow (see below for more information).

Accepted only with --generate (see below for more information).
Accepted by ofproto-trace-packet-out only. With this option, the command rejects actions that are inconsistent with the specified packet. (An example of an inconsistency is attempting to strip the VLAN tag from a packet that does not have a VLAN tag.) Open vSwitch ignores most forms of inconsistency in OpenFlow 1.0 and rejects inconsistencies in later versions of OpenFlow. The option is necessary because the command does not ordinarily imply a particular OpenFlow version. One exception is that, when actions includes an action that only OpenFlow 1.1 and later supports (such as push_vlan), --consistent is automatically enabled.
When the traced flow triggers conntrack actions, ofproto/trace will automatically trace the forked packet processing pipeline with user specified ct_state. This option sets the ct_state flags that the conntrack module will report. The flags must be a comma- or space-separated list of the following connection tracking flags:
  • trk: Include to indicate connection tracking has taken place.
  • new: Include to indicate a new flow.
  • est: Include to indicate an established flow.
  • rel: Include to indicate a related flow.
  • rpl: Include to indicate a reply flow.
  • inv: Include to indicate a connection entry in a bad state.
  • dnat: Include to indicate a packet whose destination IP address has been changed.
  • snat: Include to indicate a packet whose source IP address has been changed.
When --ct-next is unspecified, or when there are fewer --ct-next options than ct actions, the flags default to trk,new.
Most commonly, one specifies only a flow, using one of the forms above, but sometimes one might need to specify an actual packet instead of just a flow:
Some actions have side effects. For example, the normal action can update the MAC learning table, and the learn action can change OpenFlow tables. The trace commands only perform side effects when a packet is specified. If you want side effects to take place, then you must supply a packet.
(Side effects when tracing do not have external consequences. Even if a packet is specified, a trace will not output a packet or generate sFlow, NetFlow or controller events.)
Most of the time, Open vSwitch can figure out everything about the path of a packet using just the flow, but in some special circumstances it needs to look at parts of the packet that are not included in the flow. When this is the case, and you do not supply a packet, then a trace command will tell you it needs a packet.
If you wish to include a packet as part of a trace operation, there are two ways to do it:
This option, added to one of the ways to specify a flow already described, causes Open vSwitch to internally generate a packet with the flow described and then to use that packet. If your goal is to execute side effects, then --generate is the easiest way to do it, but --generate is not a good way to fill in incomplete information, because it generates packets based on only the flow information, which means that the packets really do not have any more information than the flow.
By default, for protocols that allow arbitrary L7 payloads, the generated packet has 64 bytes of payload. Use --l7-len to change the payload length, or --l7 to specify the exact contents of the payload.
This form supplies an explicit packet as a sequence of hex digits. An Ethernet frame is at least 14 bytes long, so there must be at least 28 hex digits. Obviously, it is inconvenient to type in the hex digits by hand, so the ovs-pcap(1) and ovs-tcpundump(1) utilities provide easier ways.
With this form, packet headers are extracted directly from packet, so the odp_flow or br_flow should specify only metadata. The metadata can be:
Packet QoS priority.
Mark of the packet.
Connection state of the packet.
Connection tracking zone for packet.
Connection mark of the packet.
Connection label of the packet.
The tunnel ID on which the packet arrived.
The port on which the packet arrived.
The in_port value is kernel datapath port number for the first format and OpenFlow port number for the second format. The numbering of these two types of port usually differs and there is no relationship.

Trace an unicast ICMP echo request on ingress port 1 to destination MAC 00:00:5E:00:53:01

ofproto/trace br in_port=1,icmp,icmp_type=8,\
dl_dst=00:00:5E:00:53:01


Trace an unicast ICMP echo reply on ingress port 1 to
    destination MAC 00:00:5E:00:53:01




ofproto/trace br in_port=1,icmp,icmp_type=0,\
dl_dst=00:00:5E:00:53:01




Trace an ARP request on ingress port 1




ofproto/trace br in_port=1,arp,arp_op=1




Trace an ARP reply on ingress port 1




ofproto/trace br in_port=1,arp,arp_op=2












These commands manage ovs-vswitchd's logging settings.



  

  
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 command 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 ovs-vswitchd was invoked with the
      --log-file option.

  

  
For compatibility with older versions of OVS, any is accepted as a
      word but has no effect.







  

  
Sets the log pattern for destination to pattern. Refer to
      ovs-appctl(8) for a description of the valid syntax for
      pattern.

  

  
Lists the supported logging modules and their current levels.

  

  
Lists logging patterns used for each destination.

  

  
Causes ovs-vswitchd to close its log file, if it is open. (Use
      vlog/reopen to reopen it later.)

  

  
Causes ovs-vswitchd to close its log file, if it is open, and then
      reopen it. (This is useful after rotating log files, to cause a new log
      file to be used.)

  

  
This has no effect unless ovs-vswitchd was invoked with the
      --log-file option.

  

  

    

  

  

  
By default, ovs-vswitchd limits the rate at which certain messages
      can be logged. When a message would appear more frequently than the limit,
      it is suppressed. This saves disk space, makes logs easier to read, and
      speeds up execution, but occasionally troubleshooting requires more
      detail. Therefore, vlog/disable-rate-limit allows rate limits to be
      disabled at the level of an individual log module. Specify one or more
      module names, as displayed by the vlog/list command. Specifying
      either no module names at all or the keyword any disables rate
      limits for every log module.

  

  
The vlog/enable-rate-limit command, whose syntax is the same as
      vlog/disable-rate-limit, can be used to re-enable a rate limit that
      was previously disabled.










These commands report memory usage.



  

  
Displays some basic statistics about ovs-vswitchd's memory usage.
      ovs-vswitchd also logs this information soon after startup and
      periodically as its memory consumption grows.










These commands manage ovs-vswitchd's ``coverage counters,''
    which count the number of times particular events occur during a daemon's
    runtime. In addition to these commands, ovs-vswitchd automatically
    logs coverage counter values, at INFO level, when it detects that the
    daemon's main loop takes unusually long to run.


Coverage counters are useful mainly for performance analysis and
    debugging.



  

  
Displays the averaged per-second rates for the last few seconds, the last
      minute and the last hour, and the total counts of all of the coverage
      counters.










These commands query and modify OVS tunnel components.



  

  
Adds ipv4_address/plen route to vswitchd routing table. output_bridge
      needs to be OVS bridge name. This command is useful if OVS cached routes
      does not look right.

  

  
Print all routes in OVS routing table, This includes routes cached from
      system routing table and user configured routes.

  

  
Delete ipv4_address/plen route from OVS routing table.

  

  

  

  
OVS builds ARP cache by snooping are messages. This command shows ARP
      cache table.

  

  

  

  
Adds or modifies an ARP cache entry in bridge, mapping ip to
      mac.

  

  

  

  
Flush ARP table.

  

  
Set range for UDP source port used for UDP based Tunnels. For example
      VxLAN. If case of zero arguments this command prints current range in
    use.












This section documents aspects of OpenFlow for which the OpenFlow
    specification requires documentation.






The OpenFlow specification, version 1.2, says:



  

  
Switches that implement buffering are expected to expose, through
      documentation, both the amount of available buffering, and the length of
      time before buffers may be reused.




Open vSwitch does not maintains any packet buffers.








The OpenFlow specification, version 1.4, says:



  

  
If the switch does not receive any OFPT_BUNDLE_CONTROL or
      OFPT_BUNDLE_ADD_MESSAGE message for an opened bundle_id for a switch
      defined time greater than 1s, it may send an ofp_error_msg with
      OFPET_BUNDLE_FAILED type and OFPBFC_TIMEOUT code. If the switch does not
      receive any new message in a bundle apart from echo request and replies
      for a switch defined time greater than 1s, it may send an ofp_error_msg
      with OFPET_BUNDLE_FAILED type and OFPBFC_TIMEOUT code.




Open vSwitch implements default idle bundle lifetime of 10
    seconds. (This is configurable via other-config:bundle-idle-timeout
    in the Open_vSwitch table. See ovs-vswitchd.conf.db(5) for
    details.)










We believe these limits to be accurate as of this writing. These
    limits assume the use of the Linux kernel datapath.


    
  
  • ovs-vswitchd started through ovs-ctl(8) provides a limit of
      65535 file descriptors. The limits on the number of bridges and ports is
      decided by the availability of file descriptors. With the Linux kernel
      datapath, creation of a single bridge consumes three file descriptors and
      adding a port consumes "n-handler-threads" file descriptors per
      bridge port. Performance will degrade beyond 1,024 ports per bridge due to
      fixed hash table sizing. Other platforms may have different
    limitations.
    • 
  
  • 2,048 MAC learning entries per bridge, by default. (This is configurable
      via other-config:mac-table-size in the Bridge table. See
      ovs-vswitchd.conf.db(5) for details.)
    • 
  
  • Kernel flows are limited only by memory available to the kernel.
      Performance will degrade beyond 1,048,576 kernel flows per bridge with a
      32-bit kernel, beyond 262,144 with a 64-bit kernel. (ovs-vswitchd
      should never install anywhere near that many flows.)
    • 
  
  • OpenFlow flows are limited only by available memory. Performance is linear
      in the number of unique wildcard patterns. That is, an OpenFlow table that
      contains many flows that all match on the same fields in the same way has
      a constant-time lookup, but a table that contains many flows that match on
      different fields requires lookup time linear in the number of flows.
    • 
  
  • 255 ports per bridge participating in 802.1D Spanning Tree Protocol.
    • 
  
  • 32 mirrors per bridge.
    • 
  
  • 15 bytes for the name of a port. (This is a Linux kernel limitation.)
    • 









ovs-appctl(8), ovsdb-server(1).







  

    2.10.7
    Open vSwitch