Chassis Cluster Redundancy Groups

Understanding Chassis Cluster Redundancy Group 0: Routing Engines

When you initialize a device in chassis cluster mode, the system creates a redundancy group 0 (RG0). RG0 manages the primary and failover of the Routing Engines between the cluster nodes. As with all redundancy groups, RG0 can be primary on only one node at a time.

The node on which RG0 is primary determines which Routing Engine is active in the cluster. A node is considered the primary node of the cluster when its Routing Engine is active .

The RG0 configuration includes a priority value for each node. The following priority scheme is used to determine RG0 primary. Note that the three-second interval applies when the default heartbeat-threshold and heartbeat-interval values are used.

• If one node comes online at least three seconds before the peer node, that node becomes the primary node.

• If both nodes come online at the same time, or within three seconds of each other, primary is determines as follows:

  • The node with the higher configured priority becomes the primary node.

  • If the priorities are equal (either because the same value is configured or default values are used), the node with the lower node ID becomes the primary node. In a two-node cluster, this is node 0.

The same priority scheme applies to redundancy groups numbered 1 through 128, provided that preemption is not configured.

Preemption cannot be enabled for RG0. To change the primary node for RG0, you must perform a manual failover.

Use RG0 manual failovers with caution. AnRG0 failover triggers a Routing Engine failover, during which all processes running on the current primary node are terminated and then restarted on the new primary Routing Engine. This process can result in loss of state, such as routing information, and may temporarily degrade system performance due to increased system churn.

Understanding Chassis Cluster Redundancy Groups 1 Through 128

You can configure one or more redundancy groups, numbered 1 through 128, collectively referred to as redundancy group x. The maximum number of redundancy groups that can be configured is limited by the number of redundant Ethernet interfaces supported on the device. Each redundancy group x functions as an independent unit of failover and is primary on only one node at a time.

Each redundancy group x contains one or more redundant Ethernet interfaces. A redundant Ethernet interface is a pseudo interface that consists of at least a pair of physical Gigabit Ethernet interfaces or a pair of Fast Ethernet interfaces. When a redundancy group is active on node 0, the child links of all associated redundant Ethernet interfaces on node 0 are active. When the redundancy group fails over to node 1, the child links of the associated redundant Ethernet interfaces on node 1 become active and begin forwarding traffic.

The following priority scheme determines primary for redundancy group x when preemption is not configured. If preemption is configured, the node with the higher priority becomes the primary node. Note that the three-second interval applies when the default heartbeat-threshold and heartbeat-interval values are used.

• If the node comes online at least three seconds before the peer node, that node becomes the primary node.

• If both nodes come online at the same time, or within three seconds of each other:

  • The node with the higher configured priority becomes the primary node.

  • If the priorities are equal (either because the same value is configured or default values are used), the node with the lower node IDbecomes the primary node. In a two-node cluster, this is node 0.

In a chassis cluster, you can configure multiple redundancy groups to load-share traffic across the cluster nodes. For example,

• Some redundancy groups can be primary on one node, while others are primary on the peer node.

• A redundancy group x can also be configured in a one-to-one relationship with a single redundant Ethernet interface to control the traffic path.

Traffic associated with redundancy group is processed on the node where that redundancy group is active. Because multiple redundancy groups can be configured, traffic for different redundancy groups can be processed on different nodes, depending on where each redundancy group is primary. With multiple redundancy groups, traffic can arrive on an ingress interface that belongs to one redundancy group and exit through an egress interface that belongs to another redundancy group. In some scenarios, traffic can enter the cluster through an ingress interface that belongs to one redundancy group and exit through an egress interface that belongs to another redundancy group. In this case, the ingress and egress interfaces might not be active on the same node. When this occurs, traffic is forwarded across the fabric link to the appropriate node for processing or forwarding.

When you configure a redundancy group x, you must assign a priority to each node. The node with the higher priority is selected as the primary node for that redundancy group. The primary of a redundancy group x can fail over from one node to the other. When a redundancy group x fails over, the redundant Ethernet interfaces on the new primary node become active and begin forwarding traffic.

Table 1 shows an example of redundancy group x in a chassis cluster and indicates the node on which the group is primary. It also shows the redundant Ethernet interfaces and their associated child interfaces configured for redundancy group x.

Some devices include both Gigabit Ethernet and Fast Ethernet ports.

Table 1: Example of Redundancy Groups in a Chassis Cluster

Group	Primary	Priority	Objects	Interface (Node 0)	Interface (Node 1)
Redundancy group 0	Node 0	Node 0: 254	Routing Engine on node 0	—	—
		Node 1: 2	Routing Engine on node 1	—	—
Redundancy group 1	Node 0	Node 0: 254	Redundant Ethernet interface 0	ge-1/0/0	ge-5/0/0
		Node 1: 2	Redundant Ethernet interface 1	ge-1/3/0	ge-5/3/0
Redundancy group 2	Node 1	Node 0: 2	Redundant Ethernet interface 2	ge-2/0/0	ge-6/0/0
		Node 1: 254	Redundant Ethernet interface 3	ge-2/3/0	ge-6/3/0
Redundancy group 3	Node 0	Node 0: 254	Redundant Ethernet interface 4	ge-3/0/0	ge-7/0/0
		Node 1: 2	Redundant Ethernet interface 5	ge-3/3/0	ge-7/3/0

As the example for a chassis cluster in Table 1 shows:

• The Routing Engine on node 0 is active because redundancy group 0 is primary on node 0. The Routing Engine on node 1 is passive and serves as the backup.

• Redundancy group 1 is primary on node 0. Interfaces ge-1/0/0 and ge-1/3/0 which belong to redundant Ethernet interface 0 and 1, are active and handling traffic.

• Redundancy group 2 is primary on node 1. Interfaces ge-6/0/0 and ge-6/3/0 which belong to redundant Ethernet interface 2 and 3, are active and handling traffic.

• Redundancy group 3 is primary on node 0. Interfaces ge-3/0/0 and ge-3/3/0 which belong to redundant Ethernet interface 4 and 5, are active and handling traffic.