[no] lag lag-id
config
This command enables the context in which Link Aggregation Group (LAG) attributes are defined.
A LAG groups two or more Ethernet links (ports) into one logical link. The aggregation of multiple physical links adds redundancy and improves resiliency between two network devices, and allows for load sharing.
On access, network, and hybrid ports, a LAG supports active/standby and active/active operation.
For active/standby operation, the links must be in different subgroups. Links can be on the same platform or adapter card/module or distributed over multiple components. If an active link in a LAG fails, traffic gets redistributed to the standby link.
Load sharing is supported among the active links in a LAG group.
The no form of this command deletes the LAG from the configuration. A LAG can only be deleted while it is administratively shut down. Any dependencies, such as IP interface configurations, must be removed from the LAG before it can be shut down.
no lag
the LAG identifier, expressed as a decimal integer
access
config>lag
This command enables the context to configure access parameters.
n/a
adapt-qos {link | distribute}
config>lag>access
This command specifies how active/active LAG SAP queue scheduler, SAP scheduler (H-QoS), and SAP MSS scheduler QoS rates (as specified by the SLA) are distributed to each of the active LAG ports. This command applies only to access LAGs.
link
specifies that the full QoS rates are configured on each of the active LAG links for SAP queue schedulers, SAP schedulers (H-QoS), and SAP MSS schedulers
specifies that the QoS rates are divided equally among the active LAG links for SAP queue schedulers, SAP schedulers (H-QoS), and egress MSS schedulers. For ingress MSS shaper schedulers, the distribute parameter specifies that the QoS rates are divided proportionally among the active link MDAs based on the number of active links on each MDA.
bfd
config>lag
This command creates the bfd context and enables a micro-BFD session over the associated LAG links.
n/a
family ipv4
config>lag>bfd
This command specifies the address family for the micro-BFD session over the associated LAG links.
ipv4
specifies that IPv4 encapsulation be used for the micro-BFD session
[no] bfd-on-distributing-only
config>lag>bfd>family
This command enables standardized implementation for interworking with other vendors by restricting micro-BFD sessions to links in the LACP distributing state only.
The configuration must be the same for both the local and remote LAG nodes.
The no form of this command disables restricting micro-BFD sessions; this functionality is an enhanced proprietary solution.
no bfd-on-distributing-only
local-ip-address ip-address
no local-ip-address
config>lag>bfd>family
This command is used to specify the IPv4 address of the BFD source.
The no form of this command removes the address from the configuration.
no local-ip-address
specifies the IPv4 address
max-admin-down-time [[down-interval] | infinite]
no max-admin-down-time
config>lag>bfd>family
This command specifies the maximum length of time the router continues to forward traffic over a link after the micro-BFD session has transitioned to a down state because it received an Admin Down state message from the far end. The configured time allows the administrator to disable or deprovision the micro-BFD session on the local node before forwarding is halted over the associated links.
The configured interval must be the same for both the local and remote LAG nodes.
The no form of this command removes the time interval from the configuration.
0
specifies the time, in seconds, that the router continues to forward traffic
specifies no end time to forward traffic
max-setup-time [[up-interval] | infinite]
no max-setup-time
config>lag>bfd>family
This command specifies the maximum length of time the router forwards traffic over a link that has transitioned from standby to active before the micro-BFD session is fully established (up state).
The configured interval must be the same for both the local and remote LAG nodes.
The no form of this command returns the timer value to the default (infinite), which indicates that the state of the micro-BFD session is ignored forever.
If bfd-on-distributing-only is enabled, you cannot configure the max-setup-time to a value between 0 and 3000. Similarly, if max-setup-time is set to a value between 0 and 3000, you cannot enable bfd-on-distributing-only.
infinite
specifies the time, in milliseconds, that the router forwards traffic
specifies no end time to forward traffic
multiplier multiplier
no multiplier
config>lag>bfd>family
This command specifies the detect multiplier used for a micro-BFD session over the associated LAG links. If a BFD control packet is not received within the period specified by the multiplier x receive-interval parameter of the receive-interval command, the session is declared down.
The no form of this command removes the multiplier from the configuration.
3
specifies the multiplier value
receive-interval receive-interval
no receive-interval
config>lag>bfd>family
This command specifies the receive timer used for micro-BFD sessions over the associated LAG links. The receive interval is the expected interval between received BFD control packets.
The no form of this command removes the receive timer from the configuration.
100
specifies the interval time, in milliseconds
remote-ip-address ip-address
no remote-ip-address
config>lag>bfd>family
This command is used to specify the IPv4 address of the BFD destination.
The no form of this command removes the address from the configuration.
no remote-ip-address
specifies the IPv4 address
[no] shutdown
config>lag>bfd>family
This command disables micro-BFD sessions for this address family.
The no form of this command re-enables micro-BFD sessions for this address family.
shutdown
transmit-interval transmit-interval
no transmit-interval
config>lag>bfd>family
This command specifies the transmit timer used for micro-BFD sessions over the associated LAG links. The transmit interval is the interval between transmitted BFD control packets.
The no form of this command removes the transmit timer from the configuration.
100
specifies the interval time, in milliseconds
[no] dynamic-cost
config>lag
This command enables OSPF or IS-IS costing of a LAG based on the available aggregated, operational bandwidth.
The path cost is dynamically calculated based on the interface bandwidth. OSPF path cost can be changed through the interface metric or the reference bandwidth.
If dynamic cost is configured, costing is applied based on the total number of links configured and the cost advertised is inversely proportional to the number of links available at the time. This requires that the number of links that are up exceeds the configured LAG threshold value; if the number of links that are up falls below the threshold, the configured threshold action determines whether, and at what cost, this LAG will be advertised (see port-threshold).
For example, a physical link in OSPF has a cost associated with it of 100, and the LAG consists of four physical links. The cost associated with the logical link is 25. If one link fails, the cost is automatically adjusted to 33.
If dynamic cost is not configured and OSPF auto-cost is configured, costing is applied based on the total number of links configured. This cost will remain static provided that the number of links that are up exceeds the configured LAG threshold value; if the number of links that are up falls below the threshold, the configured threshold action determines whether, and at what cost, this LAG will be advertised.
If dynamic cost is configured and OSPF auto-cost is not configured, the cost is determined by the cost configured on the OSPF metric, provided that the number of links that are up exceeds the configured LAG threshold value; if the number of links that are up falls below the threshold, the configured threshold action determines whether this LAG will be advertised.
If neither dynamic cost nor OSPF auto-cost is configured, the cost advertised is determined by the cost configured on the OSPF metric, provided that the number of links that are up exceeds the configured LAG threshold value; if the number of links that are up falls below the threshold, the configured threshold action determines whether this LAG will be advertised.
The no form of this command removes dynamic costing from the LAG.
no dynamic-cost
encap-type {dot1q | null | qinq}
no encap-type
config>lag
This command configures the encapsulation method used to distinguish customer traffic on a LAG.
The encapsulation type of a LAG must match that of its member links. If the encapsulation type of the LAG is changed, the encapsulation type of all its member links also changes.
The encapsulation type can be changed on the LAG only if there is no interface or service associated with it. If the MTU is set to a non-default value, it will be reset to the default value when the encapsulation type is changed.
The no form of this command reverts to the default encapsulation type.
null — all traffic on the link belongs to a single service or VLAN
ingress Ethernet frames carry IEEE 802.1Q tags, each tag signifying a different service
ingress Ethernet frames do not use any tags to indicate a service. As a result, only one service can be configured on a link with null encapsulation.
ingress frames carry two tags, where the outer tag is the service provider tag and the inner tag is the customer service tag as defined in 802.1ad
hold-time down hold-down-time
no hold-time
config>lag
This command specifies the delay between detecting that a LAG is down (all active ports are down) and reporting it to the upper layer protocols.
When a failure in a LAG is detected, it is immediately advertised to the rest of the system, but subsequent failures are not advertised to upper layers until the hold-time down interval has expired.
Specifying a hold time is especially useful in a 1:1 active/standby configuration because the time between detecting that the active link in the LAG is down and the time needed to activate the standby link is very short. The hold time prevents the LAG from being operationally down when switching between the active and standby link.
The no form of this command removes any hold time configured.
no hold-time
specifies, in tenths of seconds, the hold time before a failure is reported
lacp [mode] [administrative-key admin-key] [system-id system-id] [system-priority priority]
no lacp
config>lag
This command specifies the LACP mode of the LAG. By default, the LACP mode is not set.
Per the IEEE 802.1AX-2008 (IEEE 802.3ad) standard, the Link Aggregation Control Protocol (LACP) provides a standardized means for exchanging information between network devices using LAGs. LACP operates in two modes: passive and active. At least one partner must operate in active mode. For example, if the mode on the CE end is passive, the LACP mode on the 7705 SAR end must be active.
The no form of this command disables LACP.
no lacp
specifies the mode in which LACP operates
specifies an administrative key value to identify the aggregation group on each port configured to use LACP. This value should be configured only in exceptional cases. If it is not specified, a random key is assigned.
the 6-byte system identifier in hexadecimal format
the system priority
lacp-xmit-interval {slow | fast}
no lacp-xmit-interval
config>lag
This command specifies the interval signaled to the peer and tells the peer at which rate it should transmit.
The no form of this command returns the value to the default.
fast
transmits packets every 30 seconds
transmits packets every second
[no] lacp-xmit-stdby
config>lag
This command enables LACP message transmission on the standby link.
The no form of this command disables LACP message transmission on the standby link. Disable LACP message transmission on the standby link if the peer does not properly follow the IEEE 802.3ax standard regarding the LACP sync bit.
lacp-xmit-stdby
mac ieee-address
no mac
config>lag
This command assigns a specific MAC address to the LAG.
The no form of this command returns the MAC address to the default value.
a default MAC address is assigned by the system
specifies the 48-bit MAC address in the form aa:bb:cc:dd:ee:ff or aa-bb-cc-dd-ee-ff where aa, bb, cc, dd, ee, and ff are hexadecimal numbers. Allowed values are any non-broadcast, non-multicast MAC, and non-IEEE reserved MAC addresses.
mode {access | network | hybrid}
no mode
config>lag
This command configures the LAG for access, network, or hybrid mode operation.
An access port is used for customer-facing traffic on which services are configured. A SAP can only be configured on an access port or channel.
A network port participates in the service provider transport or infrastructure network when network mode is selected.
A hybrid port allows a single port to operate in both access and network modes.
The no form of this command restores the default.
network
configures the LAG for access mode operation
configures the LAG for network mode operation
configures the LAG for hybrid mode operation
port port-id [port-id] [priority priority] [sub-group sub-group-id]
no port port-id
config>lag
This command adds ports (links) to a LAG. Multiple ports can be added in one command as long as the maximum is not exceeded. The number of links in a LAG depends on the platform or adapter card/module and the Ethernet port type.
The port configuration of the first port added to the LAG is used to compare to subsequently added ports. All ports must share the same characteristics (MTU, speed, duplex, and so on) as those of the first port; otherwise, they will not be added to the LAG.
The priority parameter sets the priority of the port, which is used by LACP. The port with the highest priority is the primary port. If two or more ports share the same priority value, the port with the lowest port ID becomes the primary port.
Any configuration changes made to the primary port apply to all member ports of a LAG. For example, port hold-time and MTU settings can be modified directly on the primary port and will be applied to all member ports. Any configuration changes made to non-primary ports are rejected.
All ports in a LAG must have autonegotiation disabled or set to limited mode (recommended setting is limited). Autonegotiation can be disabled or set to limited mode using the autonegotiate command under the config>port>ethernet context.
The no form of this command removes ports from the LAG.
n/a
specifies the physical port ID in the slot/mda/port format
specifies the priority of a port
identifies a LAG subgroup. Each port in a LAG must be a member of a subgroup. If no subgroup is specified for an access port, the configuration will fail. If no subgroup is specified for a network port, the port is assigned to subgroup 1.
Subgroups should only be configured on one side of a LAG, not both. If you configure the 7705 SAR with subgroups to handle active/standby operation, the partner system should not be configured with subgroups. Only having one side perform active/standby selection guarantees a consistent selection and fast convergence. Active/standby selection is signaled through LACP from one side of the LAG to the other.
port-threshold value [action {dynamic-cost | down}]
no port-threshold
config>lag
This command sets a threshold value and controls the behavior of the LAG if the number of operational links is equal to or below the threshold value.
The no form of this command returns the value to the default.
0 action down
the number of operational links at or below which the configured action will be invoked for the LAG. When the number of operational links exceeds the port threshold value, any action taken for being below the threshold value will cease.
specifies that dynamic costing will be activated if the number of operational links is equal to or below the configured threshold value. The LAG will remain operationally up with a cost relative to the number of operational links. The link will only be regarded as operationally down when all links in the LAG are down.
specifies that the LAG will be brought operationally down if the number of operational links is equal to or below the configured threshold value. The LAG will only be regarded as up when the number of operational links exceeds the configured threshold value.
selection-criteria [best-port | highest-count | highest-weight] [slave-to-partner] [subgroup-hold-time hold-time]
no selection-criteria
config>lag
This command specifies which criteria is used to select the active subgroup (link) of a LAG. Every time the configuration of a link changes, the selection algorithm is applied.
The no form of this command returns the value to the default.
highest-count
specifies that all standby ports will have their corresponding transmitters disabled. This parameter is used in static LAG configurations.
specifies the subgroup with the highest number of eligible member links. An eligible member link of a LAG is a member that can potentially become active.
A subgroup can have a maximum of one eligible member link.
specifies the subgroup with the highest aggregate weight
selects, together with the selection criteria, the active subgroup. An eligible member link of a LAG is a member that can potentially become active. This means it is operationally up (not disabled) for use by the remote side. The slave-to-partner parameter can be used to control whether this latter condition is taken into account. The slave-to-partner parameter does not apply to static LAG configurations.
It is recommended that this parameter be set.
specifies the delay time, in tenths of a second, before switching to a newly selected active subgroup from the existing active subgroup. The timer delay applies only if the existing subgroup remains operationally up. If a value of 0 or no value is specified, the switchover occurs immediately. If a value of infinite is specified, no switchover will occur as long as the subgroup remains up; this setting can be overridden with the tools>perform>force>lag-id command. LACP must be enabled on the LAG.
standby-signaling [lacp | power-off]
no standby-signaling
config>lag
This command specifies how the state of a member port is signaled to the remote side when the status corresponding to this member port has a standby value.
The no form of this command turns off standby signaling.
specifies that lacp is active
specifies that the standby port transmitter is disabled