Juniper MX Series

Inter-Chassis Control Protocol

The Inter-Chassis Control Protocol (ICCP) is a simple and lightweight protocol that rides on top of TCP/IP that’s used to maintain state, trigger failover, and ensure the MC-LAG configuration matches between the two chassis:

MC-LAG Configuration

ICCP is able to check the following attributes to ensure that the MC-LAG configurations between chassis are sane: MC-LAG port priority, system ID, aggregator ID, and port ID offset. If a misconfiguration is detected, MC-LAG will not come up properly; this is very helpful in quickly identifying and isolating operational MC-LAG problems.

State Information

In order to provide a transparent IEEE 802.3ad interface to a downstream CE, there are runtime objects that need to be synchronized between the two separate chassis: IGMP and DHCP snooping specific to interfaces participating in MC-LAG, MAC addresses that are learned or installed between the different chassis, and MC-LAG interfaces and their operational status.

Status Information

If the MC-LAG is running in an Active-Standby state, ICCP will need to keep track of which chassis is currently active versus in standby state.

Change Request

As the topology changes during a network event such as an interface doing down, ICCP will need to react accordingly; an example change request would be changing the MC-LAG state of Active from PE1 to PE2.

The ICCP protocol is required when creating a MC-LAG configuration. Keep in mind that ICCP is just a simple control protocol and doesn’t ...