Observe the following general rules and conditions when configuring LAGs.
All ports (links) in a LAG must share the same characteristics (speed, duplex, hold-timer, and so on). The port characteristics are inherited from the primary port.
Autonegotiation must be disabled or set to limited mode for ports in a LAG, in order to guarantee a specific port speed.
Ports in a LAG must be configured as full duplex.
Ports in a LAG must be configured with the same encapsulation value.
LAG is supported on Ethernet access, network, and hybrid ports.
On access ports, the links must be distributed over two different adapter cards or different MDAs on the 7705 SAR-X, in order to minimize the effect of an adapter card failure on the LAG.
On network ports, the links can be on the same platform or adapter card/module or distributed over multiple components.
A LAG can be in active/active mode or in active/standby mode for access, network, or hybrid mode. Active/standby mode is a subset of active/active mode if subgroups are enabled.
By default, LACP is disabled. LACP operates in two modes: passive and active. If the mode on the CE end is passive, the LACP mode on the 7705 SAR end must be active.
Each link in a LAG must be a member of a subgroup. On access, network or hybrid ports, a LAG can have a maximum of four subgroups and a subgroup can have links up to the maximum number supported on the LAG. LAG is active/active if there is only one sub-group. LAG is active/standby if there is more than one subgroup.
The port with the highest priority is the primary port. If multiple ports have the same priority, the port with the lowest port ID becomes the primary port.
A port on standby can be replaced while the active port in the LAG is operational.
When one port is on a Gen-2 Ethernet adapter card and the other port is on a Gen-3 Ethernet adapter card, mix-and-match traffic management occurs. The LAG SAP uses a generic QoS configuration where scheduler-mode, agg-rate, and cir-rate are configured for the SAP, but only those applicable parameters needed by the active adapter card are used to set the QoS values of the active port. See LAG Support on Mixed-Generation Hardware for details. See Table: Ethernet Adapter Card, Module, and Platform Generations for a list of adapter card generations.
The primary port configuration settings are applied to both the primary and secondary LAG ports. Therefore, in order to support unshaped SAPs when the primary port is a Gen-3-based port and the secondary port is a Gen-2-based port, configuring the unshaped-sap-cir on the Gen-3-based port is allowed, even though it does not apply to the Gen-3-based port. This is because unshaped-sap-cir is needed by the (secondary) Gen-2-based port when it becomes the active port. The full command is config>port>ethernet>access> egress>unshaped-sap-cir cir-rate.
Additional general rules for LAG configuration are as follows.
Most port features (port commands) can only be configured on the primary member port. The configuration, or any change to the configuration, is automatically propagated to any remaining ports within the same LAG. Operators cannot modify the configurations on non-primary ports.
When adding the first port member to a LAG group, its port configuration becomes the configuration of the LAG group.
Once a LAG group has been created, new ports can be added to the LAG group only if their port configurations match with the LAG group configurations inherited from the group’s existing primary port. A newly added port may become the primary port or a non-primary port, depending on the setting of the LAG selection criteria, priorities, and so on.
Not all configurations follow the conventions above. Some exceptions include the commands loopback (internal or line), cfm-loopback, mac, lldp, dot1x, and efm-oam.
At boot-up, port configuration is applied before LAG configuration is applied. Therefore, configuration values are allowed or prohibited for both a standalone port and a port attached to a LAG group; otherwise, a database restore or exec command will fail. For example, if the scheduler mode profile command option is supported on a Gen-2 port that is the primary member of a LAG, operators are allowed to change scheduler-mode on other member links, including Gen-3 ports. If the scheduler mode profile command option is blocked for a standalone Gen-2 port, then after issuing an admin>save command followed by a reboot command, the node will fail to reload the database file.
The following CLI syntax shows an example of configuring LAG parameters:
config# lag 2
config>lag# description LAG2
config>lag# port 1/4/3 priority 200 sub-group 1
config>lag# port 1/5/3 priority 100 sub-group 2
config>lag# lacp active administrative-key 32768
The following example displays a LAG configuration:
NOK-B>config>lag# info detail
----------------------------------------------
shutdown
description "LAG2"
port 1/4/3 priority 200 sub-group 1
port 1/5/3 priority 100 sub-group 2
lacp active administrative-key 32768
no mac
mode access
encap-type null
port-threshold 0
lacp-xmit-interval fast
lacp-xmit-stdby
no selection-criteria
no hold-time
standby-signaling lacp
----------------------------------------------
NOK-B>config>lag#