KB40808 - Cluster VIP fails over occur when active node becomes unreachable, the fails over again when the node becomes reachable again

This article describes an issue where the cluster VIP fails over twice when a node becomes unreachable and reachable again.

In a active / passive cluster, the cluster VIP may fail over twice under the following conditions:

• Active node becomes unreachable (due to multiple factors), then the cluster vip will fail over the passive node
• Once the node becomes reachable, the cluster vip will fail over back to the original node

This issue is only applicable to the following releases:

• 8.2R1 to 8.2R2
• 8.1R1 to 8.1R8

This issue occurs due to a behavior change of vip ownership starting in 8.1 release.

To resolve this issue, the recommendation is to upgrade to 8.2R3+ or 8.1R9+. A code change was made to ensure the behavior is similar to 8.0RX and below where the VIP remains on the existing node once the problematic node becomes reachable again.