Circuit-ID Tunnels—Rules and Notes

Keep in mind the following rules and notes when working with Circuit-ID tunnel encapsulation and decapsulation:

■

Unique Circuit IDs per cluster (prior to 6.12): Maximum 512 unique circuit IDs per cluster or standalone node for encapsulation and decapsulation.

■

Fabric maps per cluster-pair (non-shared mode): Maximum 1024 fabric maps between any two given clusters or nodes.

■

Fabric maps per circuit GigaStream (decapsulation): A single circuit GigaStream can source and decapsulation traffic from up to 1024 fabric maps.

■

If a network port receives a double-tagged packet that is encapsulated with a circuit-ID, the five tuple hashing will not work only in the second cluster, that is the cluster in the decapsulation side over stack GigaStream or tool gigastream. Hence, traffic cannot be filtered using the IP/L4 parameters. After decapsulation, Flow Mapping® filters the traffic based on circuit-ID.

■

It is not supported for inline scenarios.

Keep in mind the following rules and notes when working with Circuit-ID tunnel encapsulation:

■

Circuit-ID tunnel encapsulation is not supported on Pass All maps.

■

Port filter configured on circuit port for VLAN pass/drop will try to match the encap circuit-id instead of packet outer VLAN.

Keep in mind the following rules and notes when working with Circuit-ID tunnel decapsulation:

■

A maximum of 512 circuit-ID tunnels can be created for decapsulation.

■

Circuit-ID tunnel decapsulation is not supported on Pass All and Shared Collector maps.

■

A circuit-ID must be paired with a circuit port, only in one circuit-ID tunnel.

Keep in mind the following rules and notes when working with Fabric map circuit ID allocation:

■

Circuit ID is used internally to pass traffic from one cluster to another. You can configure Circuit ID ranges from 2 to 4000, or set your own custom range, with each topology able to reuse ID ranges (e.g., 2-513) across multiple topologies.

■

Circuit ID allocation is managed globally, beginning from the lower limit of the defined range. This ensures that IDs are allocated efficiently across different topologies. It is allocated automatically.