Contents

Configuring failover groups· 1

About failover groups· 1

Failover group operating mechanism·· 1

Failover group types· 1

Backup types· 1

Restrictions and guidelines: failover group configuration· 2

Configuring a failover group· 2

Display and maintenance commands for failover groups· 2

Configuring failover groups

About failover groups

A failover group backs up services (such as NAT) between two CPUs to ensure service continuity. A CPU is called a node in the failover group feature.

Failover group operating mechanism

A failover group contains one primary node and one secondary node. When a failover group is used by a service module, the primary node processes services and backs up service data to the secondary node. The secondary node does not process services, but it takes over the services when the primary node fails. When the primary node is recovered, traffic is switched back to the primary node.

For more information about the service modules that can use a failover group, see the configuration guides for specific modules.

Failover group types

A failover group can be automatically created by the system or manually configured.

Automatic failover group

When a security engine is installed, the system automatically creates a failover group for the security engine. The automatic failover group for a security engine uses one CPU of the security engine as the primary node. Typically, an automatic failover group does not have a secondary node and does not back up services. If multiple security engines are available on the device, you can specify a security engine for a specific service. When the security engine is removed, the system automatically allocates services evenly among the other security engines in the same security engine group. For more information about security engine groups, see the context feature in Virtual Technologies Configuration Guide.

Manual failover group

A manual failover group is manually created, to which you can bind one primary node and one secondary node.

Backup types

You can deployment service failover by using one, two, or multiple failover groups, as follows:

·     One failover group for two nodes—Assign the two nodes to one failover group as the primary and secondary nodes. In this deployment, only one node processes services.

·     Two failover groups for two nodes—Assign the primary node and the secondary node in one failover group to the other failover group as the secondary node and the primary node. In this deployment, both nodes process services and back up service data to each other.

·     Multiple failover groups for more than two nodes—Include the following deployments:

¡     Centralized backup, in which one secondary node backs up multiple primary nodes. In this deployment, configure different primary nodes but the same secondary node in different failover groups.

¡     Load-balanced backup, in which a node uses one node for backup and provides backup for another node. In this deployment, you must make sure each failover group has a unique pair of nodes, avoiding using any two nodes to back up each other. For example, to deploy load-balanced backup across three nodes (nodes 1, 2, and 3), create three failover groups. Then, configure nodes 1 and 2, nodes 2 and 3, and nodes 3 and 1 as the primary and secondary nodes in the three failover groups.

Restrictions and guidelines: failover group configuration

As a best practice, assign two CPUs with similar performance on different cards to a failover group. This operation ensures service continuity after a primary and secondary node switchover.

Configuring a failover group

1.     Enter system view.

system-view

2.     Create a failover group and enter its view.

failover group group-name

By default, one or multiple automatic failover groups prefixed with AutoBackup exist, and no manual failover groups exist.

3.     Assign a node to the failover group.

bind chassis chassis-number slot slot-number { primary | secondary }

By default, a failover group does not contain nodes.

The primary and secondary nodes in a failover group cannot be the same one. Different failover groups cannot share the same primary node.

Display and maintenance commands for failover groups

Execute display commands in any view.

‌