STP and Hitless Failover—SummitStack Only

When you are using redundancy in a SummitStack, one node assumes the role of primary and the other node assumes the role of backup. The primary executes the switch‘s management functions, and the backup acts in a standby role. Hitless failover transfers switch management control from the primary to the backup and maintains the state of STP. STP supports hitless failover. You do not explicitly configure hitless failover support; rather, if you have two nodes installed, hitless failover is available.

Note

Not all platforms support hitless failover in the same software release. To verify if the software version you are running supports hitless failover, see the following table in Managing the Switch. For more information about protocol and platform support for hitless failover, see Understanding Hitless Failover Support.

To support hitless failover, the primary node replicates STP BPDUs to the backup, which allows the nodes to run STP in parallel. Although both primary and backup node receive STP BPDUs, only the primary transmits STP BPDUs to neighboring switches and participates in STP.

Note

Before initiating failover, review the section Synchronizing Nodes to confirm that both primary and backup nodes are running software that supports the synchronize command.

To initiate hitless failover on a network that uses STP:

Confirm that the nodes are synchronized and have identical software and switch configurations using the command:
show switch {detail}

The output displays the status of the primary and backup nodes, with the primary node showing MASTER and the backup node showing BACKUP (InSync).

If the primary and backup nodes are not synchronized and both nodes are running a version of ExtremeXOS that supports synchronization, proceed to 2.

If the primary and backup nodes are synchronized, proceed to 3.
If the primary and backup nodes are not synchronized, use the synchronize command to replicate all saved images and configurations from the primary to the backup.
After you confirm the nodes are synchronized, proceed to 3.
If the nodes are synchronized, use the run failover (formerly run msm-failover) command to initiate failover.

For more detailed information about verifying the status of the primary and backup nodes, and system redundancy, see Understanding System Redundancy. For more information about hitless failover, see Understanding Hitless Failover Support.