X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=vswitchd%2Fvswitch.xml;h=78594e75e4f787de80783c8fe7031378e2ce77eb;hb=003ce655b7116d18c86a74c50391e54990346931;hp=03c8539d0a448e31a5594f6173c00ae6f514b278;hpb=c75b7e39d973cc9f4869c84d48eeb3b66afb2971;p=sliver-openvswitch.git
diff --git a/vswitchd/vswitch.xml b/vswitchd/vswitch.xml
index 03c8539d0..78594e75e 100644
--- a/vswitchd/vswitch.xml
+++ b/vswitchd/vswitch.xml
@@ -71,6 +71,106 @@
The Citrix XenServer universally unique identifier for the physical
host as displayed by
+ When
+ This option allows for improvement. When
+ Thus, with this option, the procedure for a hot-upgrade of
+
+ The
+ The maximum
+ number of flows allowed in the datapath flow table. Internally OVS
+ will choose a flow limit which will likely be lower than this number,
+ based on real time network conditions.
+
+ The default is 200000.
+
+ Specifies the number of threads for software datapaths to use for
+ handling new flows. The default the number of online CPU cores minus
+ the number of revalidators.
+
+ This configuration is per datapath. If you have more than one
+ software datapath (e.g. some
+ Specifies the number of threads for software datapaths to use for
+ revalidating flows in the datapath. Typically, there is a direct
+ correlation between the number of revalidator threads, and the number
+ of flows allowed in the datapath. The default is the number of cpu
+ cores divided by four plus one. If
+ This configuration is per datapath. If you have more than one
+ software datapath (e.g. some xe host-list
.
+
+ ovs-vswitchd
starts up, it has an empty flow table
+ and therefore it handles all arriving packets in its default fashion
+ according to its configuration, by dropping them or sending them to
+ an OpenFlow controller or switching them as a standalone switch.
+ This behavior is ordinarily desirable. However, if
+ ovs-vswitchd
is restarting as part of a ``hot-upgrade,''
+ then this leads to a relatively long period during which packets are
+ mishandled.
+ ovs-vswitchd
+ starts with this value set as true
, it will neither
+ flush or expire previously set datapath flows nor will it send and
+ receive any packets to or from the datapath. When this value is
+ later set to false
, ovs-vswitchd
will
+ start receiving packets from the datapath and re-setup the flows.
+ ovs-vswitchd
becomes roughly the following:
+
+
+ ovs-vswitchd
.
+ true
.
+ ovs-vswitchd
.
+ ovs-ofctl
(or some other program, such as an
+ OpenFlow controller) to restore the OpenFlow flow table
+ to the desired state.
+ false
(or remove it entirely from the database).
+ ovs-ctl
's ``restart'' and ``force-reload-kmod''
+ functions use the above config option during hot upgrades.
+ system
bridges and some
+ netdev
bridges), then the total number of threads is
+ n-handler-threads
times the number of software
+ datapaths.
+ n-handler-threads
is
+ set, the default changes to the number of cpu cores minus the number
+ of handler threads.
+ system
bridges and some
+ netdev
bridges), then the total number of threads is
+ n-handler-threads
times the number of software
+ datapaths.
+
If this value is unset, the default is implementation-specific.
+
+ The default is standalone
if the value is unset, but
+ future versions of Open vSwitch may change the default.
+
+ The standalone
mode can create forwarding loops on a
+ bridge that has more than one uplink port unless STP is enabled. To
+ avoid loops on such a bridge, configure secure
mode or
+ enable STP (see ).
+
When more than one controller is configured, is considered only when none of the configured controllers can be contacted.
@@ -430,6 +543,12 @@ value. May not be all-zero. +switch3 in room 3120
.
+ true
, disable in-band control on the bridge
@@ -444,6 +563,25 @@
QoS configured, or if the port does not have a queue with the specified
ID, the default queue is used instead.
+ List of OpenFlow protocols that may be used when negotiating a
+ connection with a controller. A default value of
+ OpenFlow10
will be used if this column is empty.
+
+ The current implementation of OpenFlow 1.4 support is not safe:
+ ovs-vswitchd
will abort when certain unimplemented
+ features are tested. Thus, for now it is suitable only for
+ experimental use. For this reason, OpenFlow 1.4 is supported only
+ if, in addition to specifying OpenFlow14
in this field,
+ ovs-vswitchd
is invoked with the
+ --enable-of14
option. (When support becomes safe, this
+ option will be removed.)
+
- A number of flows as a nonnegative integer. This sets number of - flows at which eviction from the kernel flow table will be triggered. - If there are a large number of flows then increasing this value to - around the number of flows present can result in reduced CPU usage - and packet loss. -
-- The default is 1000. Values below 100 will be rounded up to 100. -
-true
to enable.
+
+ The following destination MAC addresss will not be forwarded when this
+ option is enabled.
+ 01:80:c2:00:00:00
01:80:c2:00:00:01
01:80:c2:00:00:0x
00:e0:2b:00:00:00
00:e0:2b:00:00:04
and 00:e0:2b:00:00:06
+ 01:00:0c:cc:cc:cc
01:00:0c:cc:cc:cd
01:00:0c:cd:cd:cd
01:00:0c:00:00:00
01:00:0c:cc:cc:cx
+ The maximum number of MAC addresses to learn. The default is + currently 2048. The value, if specified, is forced into a reasonable + range, currently 10 to 1,000,000. +
+A port that has more than one interface is a ``bonded port.'' Bonding - allows for load balancing and fail-over. Some kinds of bonding will - work with any kind of upstream switch:
+ allows for load balancing and fail-over. + ++ The following types of bonding will work with any kind of upstream + switch. On the upstream switch, do not configure the interfaces as a + bond: +
balance-slb
active-backup
The following modes require the upstream switch to support 802.3ad with
- successful LACP negotiation:
+ successful LACP negotiation. If LACP negotiation fails and
+ other-config:lacp-fallback-ab is true, then active-backup
+ mode is used:
stable
Attempts to always assign a given flow to the same slave
- consistently. In an effort to maintain stability, no load
- balancing is done. Uses a similar hashing strategy to
- balance-tcp
, always taking into account L3 and L4
- fields even if LACP negotiations are unsuccessful.
Slave selection decisions are made based on if set. Otherwise,
- OpenFlow port number is used. Decisions are consistent across all
- ovs-vswitchd
instances with equivalent
-
- values.
These columns apply only to bonded ports. Their values are
@@ -841,7 +1007,7 @@
- The number of milliseconds for which carrier must stay up on an
+ The number of milliseconds for which the link must stay up on an
interface before the interface is considered to be up. Specify
The LACP timing which should be used on this .
- Possible values are
- Users may manually set a heartbeat transmission rate to increase
- the fault detection speed further. When manually set, OVS expects
- the partner switch to be configured with the same transmission
- rate. Manually setting 0
to enable the interface immediately.
0
to disable the interface immediately.
off
if unset.
+ disabled, unless other-config:lacp-fallback-ab is set to true.
+ Defaults to off
if unset.
fast
, slow
and a
- positive number of milliseconds. By default slow
is
- used. When configured to be fast
LACP heartbeats are
- requested at a rate of once per second causing connectivity
- problems to be detected more quickly. In slow
mode,
- heartbeats are requested at a rate of once every 30 seconds.
+ By default slow
is used. When configured to be
+ fast
LACP heartbeats are requested at a rate of once
+ per second causing connectivity problems to be detected more
+ quickly. In slow
mode, heartbeats are requested at a
+ rate of once every 30 seconds.
lacp-time
to something other
- than fast
or slow
is not supported by the
- LACP specification.
+ Determines the behavior of openvswitch bond in LACP mode. If
+ the partner switch does not support LACP, setting this option
+ to true
allows openvswitch to fallback to
+ active-backup. If the option is set to false
, the
+ bond will be disabled. In both the cases, once the partner switch
+ is configured to LACP mode, the bond will use LACP.
false
.
-
These settings control behavior when a bond is in
- balance-slb
mode, regardless of whether the bond was
- intentionally configured in SLB mode or it fell back to SLB mode
- because LACP negotiation failed.
+ balance-slb
or balance-tcp
mode.
Ethernet address to set for this interface. If unset then the default MAC address is used:
@@ -1109,19 +1279,59 @@ address.OpenFlow port number for this interface. Unlike most columns, this - column's value should be set only by Open vSwitch itself. Other - clients should set this column to an empty set (the default) when - creating an .
-Open vSwitch populates this column when the port number becomes - known. If the interface is successfully added, - will be set to a number between 1 and 65535 - (generally either in the range 1 to 65279, inclusive, or 65534, the - port number for the OpenFlow ``local port''). If the interface - cannot be added then Open vSwitch sets this column - to -1.
-+ When a client adds a new interface, Open vSwitch chooses an OpenFlow + port number for the new port. If the client that adds the port fills + in , then Open vSwitch tries to use its + value as the OpenFlow port number. Otherwise, or if the requested + port number is already in use or cannot be used for another reason, + Open vSwitch automatically assigns a free port number. Regardless of + how the port number was obtained, Open vSwitch then reports in the port number actually assigned. +
+ ++ Open vSwitch limits the port numbers that it automatically assigns to + the range 1 through 32,767, inclusive. Controllers therefore have + free use of ports 32,768 and up. +
+ ++ OpenFlow port number for this interface. Open vSwitch sets this + column's value, so other clients should treat it as read-only. +
+
+ The OpenFlow ``local'' port (OFPP_LOCAL
) is 65,534.
+ The other valid port numbers are in the range 1 to 65,279,
+ inclusive. Value -1 indicates an error adding the interface.
+
+ Requested OpenFlow port number for this interface. +
+ ++ A client should ideally set this column's value in the same + database transaction that it uses to create the interface. Open + vSwitch version 2.1 and later will honor a later request for a + specific port number, althuogh it might confuse some controllers: + OpenFlow does not have a way to announce a port number change, so + Open vSwitch represents it over OpenFlow as a port deletion + followed immediately by a port addition. +
+ ++ If is set or changed to some other + port's automatically assigned port number, Open vSwitch chooses a + new port number for the latter port. +
+gre
ipsec_gre
capwap
gre64
ipsec_gre64
vxlan
+ An Ethernet tunnel over the experimental, UDP-based VXLAN
+ protocol described at
+ http://tools.ietf.org/html/draft-mahalingam-dutt-dcops-vxlan-03
.
+
+ Open vSwitch uses UDP destination port 4789. The source port used for + VXLAN traffic varies on a per-flow basis and is in the ephemeral port + range. +
+lisp
+ A layer 3 tunnel over the experimental, UDP-based Locator/ID + Separation Protocol (RFC 6830). +
++ Only IPv4 and IPv6 packets are supported by the protocol, and + they are sent and received without an Ethernet header. Traffic + to/from LISP ports is expected to be configured explicitly, and + the ports are not intended to participate in learning based + switching. As such, they are always excluded from packet + flooding. +
patch
null
These options apply to interfaces with of
- gre
, ipsec_gre
, and capwap
.
+ gre
, ipsec_gre
, gre64
,
+ ipsec_gre64
, vxlan
, and lisp
.
@@ -1201,22 +1447,67 @@
- Required. The tunnel endpoint. Unicast and multicast endpoints are - both supported. -
+Required. The remote tunnel endpoint, one of:
+ +192.168.0.123
.
+ Only unicast endpoints are supported.
+ flow
. The tunnel accepts packets from any
+ remote tunnel endpoint. To process only packets from a specific
+ remote tunnel endpoint, the flow entries may match on the
+ tun_src
field. When sending packets to a
+ remote_ip=flow
tunnel, the flow actions must
+ explicitly set the tun_dst
field to the IP address of
+ the desired remote tunnel endpoint, e.g. with a
+ set_field
action.
+
- When a multicast endpoint is specified, a routing table lookup occurs
- only when the tunnel is created. Following a routing change, delete
- and then re-create the tunnel to force a new routing table lookup.
+ The remote tunnel endpoint for any packet received from a tunnel
+ is available in the tun_src
field for matching in the
+ flow table.
+ Optional. The tunnel destination IP that received packets must + match. Default is to match all addresses. If specified, may be one + of: +
+ +192.168.12.3
.
+ flow
. The tunnel accepts packets sent to any
+ of the local IP addresses of the system running OVS. To process
+ only packets sent to a specific IP address, the flow entries may
+ match on the tun_dst
field. When sending packets to a
+ local_ip=flow
tunnel, the flow actions may
+ explicitly set the tun_src
field to the desired IP
+ address, e.g. with a set_field
action. However, while
+ routing the tunneled packet out, the local system may override the
+ specified address with the local IP address configured for the
+ outgoing system interface.
+
+
+ This option is valid only for tunnels also configured with the
+ remote_ip=flow
option.
+
+ The tunnel destination IP address for any packet received from a
+ tunnel is available in the tun_dst
field for matching in
+ the flow table.
+
flow
. The tunnel accepts packets with any
@@ -1255,8 +1547,9 @@
key="out_key"/> at all.
flow
. Packets sent through the tunnel will
@@ -1275,7 +1568,8 @@
inherit
, in which case
+ packet. ToS is interpreted as DSCP and ECN bits, ECN part must be
+ zero. It may also be the word inherit
, in which case
the ToS will be copied from the inner packet if it is IPv4 or IPv6
(otherwise it will be 0). The ECN fields are always inherited.
Default is 0.
@@ -1288,48 +1582,13 @@
system default, typically 64). Default is the system default TTL.
true
to
- enable.
- df_inherit
option is not set, or if
- the encapsulated packet is not IP. Default is enabled; set to
- false
to disable.
- false
to disable.
+ Optional. If enabled, the Don't Fragment bit will be set on tunnel
+ outer headers to allow path MTU discovery. Default is enabled; set
+ to false
to disable.
- Only gre
interfaces support these options.
-
iptables
) and it may be useful to disable it if these
- features are required or as a debugging measure. Default is enabled,
- set to false
to disable.
-
Only gre
and ipsec_gre
interfaces support
@@ -1489,15 +1748,15 @@
gre
or capwap
.
+ gre
.
+ BFD, defined in RFC 5880 and RFC 5881, allows point-to-point + detection of connectivity failures by occasional transmission of + BFD control messages. Open vSwitch implements BFD to serve + as a more popular and standards compliant alternative to CFM. +
+ ++ BFD operates by regularly transmitting BFD control messages at a rate + negotiated independently in each direction. Each endpoint specifies + the rate at which it expects to receive control messages, and the rate + at which it is willing to transmit them. Open vSwitch uses a detection + multiplier of three, meaning that an endpoint signals a connectivity + fault if three consecutive BFD control messages fail to arrive. In the + case of a unidirectional connectivity issue, the system not receiving + BFD control messages signals the problem to its peer in the messages it + transmits. +
+ ++ The Open vSwitch implementation of BFD aims to comply faithfully + with RFC 5880 requirements. Open vSwitch does not implement the + optional Authentication or ``Echo Mode'' features. +
+ ++ A controller sets up key-value pairs in the + column to enable and configure BFD. +
+ +1000
.
+ 100
.
+ 00:23:20:00:00:01
.
+ 169.254.1.0
.
+ 169.254.1.1
.
+ + The switch sets key-value pairs in the + column to report the status of BFD on this interface. When BFD is + not enabled, with , the switch clears + all key-value pairs from . +
+ +UP
.
+ UP
, and the remote
+ system isn't signaling a problem such as concatenated path down.
+ 802.1ag Connectivity Fault Management (CFM) allows a group of @@ -1663,12 +2073,30 @@ faulted otherwise.
+
+ When operating over tunnels which have no in_key
, or an
+ in_key
of flow
. CFM will only accept CCMs
+ with a tunnel key of zero.
+
+ A Maintenance Point ID (MPID) uniquely identifies each endpoint + within a Maintenance Association. The MPID is used to identify this + endpoint to other Maintenance Points in the MA. Each end of a link + being monitored should have a different MPID. Must be configured to + enable CFM on this . +
++ According to the 802.1ag specification, MPIDs can only range between + [1, 8191]. However, extended mode (see ) supports eight byte MPIDs. +
+ovs-appctl
command.
When in extended mode, indicates the operational state of the
+ remote endpoint as either up
or down
. See
+ .
+
Indicates the health of the interface as a percentage of CCM frames
@@ -1757,9 +2197,21 @@
+ The interval, in milliseconds, between transmissions of CFM
+ heartbeats. Three missed heartbeat receptions indicate a
+ connectivity fault.
+
+ In standard operation only intervals of 3, 10, 100, 1,000, 10,000,
+ 60,000, or 600,000 ms are supported. Other values will be rounded
+ down to the nearest value on the list. Extended mode (see ) supports any interval up
+ to 65,535 ms. In either mode, the default is 1000 ms.
+ We do not recommend using intervals less than 100 ms.
+ When
+ Demand mode has a couple of caveats:
+ false
.
+ compatibility with 802.1ag compliant implementations. And extended
+ mode allows eight byte MPIDs. Defaults to false
.
true
, and
+ is true, the CFM
+ module operates in demand mode. When in demand mode, traffic
+ received on the is used to indicate
+ liveness. CCMs are still transmitted and received, but if the
+ is receiving traffic, their absence does not
+ cause a connectivity fault.
+
+
+
down
, the CFM module marks all CCMs it generates as
@@ -1794,7 +2284,7 @@
stable
bond mode to make slave
- selection decisions. Allocating values consistently across interfaces
- participating in a bond will guarantee consistent slave selection
- decisions across ovs-vswitchd
instances when using
- stable
bonding mode.
-
+ Hypervisors may sometimes have more than one interface associated
+ with a given , only one of
+ which is actually in use at a given time. For example, in some
+ circumstances XenServer has both a ``tap'' and a ``vif'' interface
+ for a single , but only
+ uses one of them at a time. A hypervisor that behaves this way must
+ mark the currently in use interface active
and the
+ others inactive
. A hypervisor that never has more than
+ one interface for a given
+ may mark that interface active
or omit entirely.
+
+ During VM migration, a given might transiently be marked active
on
+ two different hypervisors. That is, active
means that
+ this is the active
+ instance within a single hypervisor, not in a broader scope.
+ There is one exception: some hypervisors support ``migration'' from a
+ given hypervisor to itself (most often for test purposes). During
+ such a ``migration,'' two instances of a single might both be briefly marked
+ active
on a single hypervisor.
+
The eviction process only considers flows that have an idle timeout or a hard timeout. That is, eviction never deletes permanent flows. - (Permanent flows do count against . + (Permanent flows do count against .)
@@ -2055,6 +2566,75 @@ column has no effect.
+ This string set specifies which fields should be used for + address prefix tracking. Prefix tracking allows the + classifier to skip rules with longer than necessary prefixes, + resulting in better wildcarding for datapath flows. +
++ Prefix tracking may be beneficial when a flow table contains + matches on IP address fields with different prefix lengths. + For example, when a flow table contains IP address matches on + both full addresses and proper prefixes, the full address + matches will typically cause the datapath flow to un-wildcard + the whole address field (depending on flow entry priorities). + In this case each packet with a different address gets handed + to the userspace for flow processing and generates its own + datapath flow. With prefix tracking enabled for the address + field in question packets with addresses matching shorter + prefixes would generate datapath flows where the irrelevant + address bits are wildcarded, allowing the same datapath flow + to handle all the packets within the prefix in question. In + this case many userspace upcalls can be avoided and the + overall performance can be better. +
++ This is a performance optimization only, so packets will + receive the same treatment with or without prefix tracking. +
+
+ The supported fields are: tun_id
,
+ tun_src
, tun_dst
,
+ nw_src
, nw_dst
(or aliases
+ ip_src
and ip_dst
),
+ ipv6_src
, and ipv6_dst
. (Using this
+ feature for tun_id
would only make sense if the
+ tunnel IDs have prefix structure similar to IP addresses.)
+
+ For example, prefixes=ip_dst,ip_src
instructs the
+ flow classifier to track the IP destination and source
+ addresses used by the rules in this specific flow table. To
+ set the prefix fields, the flow table record needs to exist:
+
ovs-vsctl set Bridge br0 flow_tables:0=@N1 -- --id=@N1 create Flow_Table name=table0
ovs-vsctl set Flow_Table table0 prefixes=ip_dst,ip_src
+ There is a maximum number of fields that can be enabled for any + one flow table. Currently this limit is 3. +
+Common
+ Columns
at the beginning of this document.
+
+