M-LAG also applies to scenarios where a server is dual-homed to two access switches with two NICs working in active-active mode. Scheme 2: Access switches set up an M-LAG system.Additionally, M-LAG devices can be upgraded independently, facilitating device maintenance. M-LAG offers higher reliability than stack. M-LAG devices have independent control and management planes and use the same protocol for information synchronization. M-LAG provides a clearer logical topology and higher link efficiency than the traditional STP loop prevention mechanism. The ports on the two aggregation switches connected to the same access switch set up an inter-chassis Eth-Trunk.įigure 5-4 Deploying M-LAG on aggregation switches Two aggregation switches are configured to set up an M-LAG system and the link between the switches is configured as a peer-link.

M-LAG enables a loop-free logical network between aggregation and access switches, making STP deployment unnecessary.

Scheme 1: Aggregation switches set up an M-LAG system.

The two M-LAG devices are independent, so the in-band or out-of-band IP reachability detection can be used for DAD, without requiring additional cabling. M-LAG supports dual-active detection (DAD). In addition, M-LAG supports preferential forwarding of local traffic, minimizing east-to-west traffic between the two switches.

In the Layer 3 logical topology, the two switches are two independent devices, can be managed by independent NMSs, and function as independent OSPF nodes.

A peer-link is configured between the two switches to exchange M-LAG packets and forward horizontal service traffic between the switches. In Figure 5-3, the two M-LAG switches provide M-LAG interfaces for Layer 2 service access. Figure 5-3 M-LAG-based physical topology and Layer 2 as well as Layer 3 logical topologies

Free software lag switch