Duplicate VIP or SCAN VIP or SCAN Listener Found on More than 1 Node in a RAC Cluster (Doc ID 1575272.1)

Last updated on SEPTEMBER 16, 2021

Applies to:

Symptoms

Duplicate IP for host VIP or SCAN VIP or SCAN listener found on more than 1 node in the cluster after a failover or CRS restart. In other words, the IP for the host VIP or SCAN VIP is not unplumbed from the network of its original node, causing client connection problem when that particular VIP or SCAN VIP is used.

In this sample, it is 2 node RAC cluster,
16:21:51 CSSD on node <nodename2> shutdown by user.
CRS could not start properly due to ASM issue until
17:48:27 CSSD Reconfiguration complete. Active nodes are <nodename1> <nodename2> .
17:57:03 CRSD started on node <nodename2>
18:27:29 CSSD on node <nodename1> shutdown by user
18:28:40 CSSD Reconfiguration complete. Active nodes are <nodename1> <nodename2>
18:29:03 CRSD started on node <nodename1>

After this, <nodename2>-vip shows ONLINE on node 2, but its IP is binded on both nodes, its IP was not unplumbed from node 1:

<nodename1>-vip is 10.xx.xx.19
<nodename2>-vip is 10.xx.xx.20

Node 1 racnode01  ifconfig output

Above sample show host VIP <nodename2>-vip 10.xx.xx.20 is plumbed on both nodes. This problem could also happen to SCAN VIP. The scenario leading to this situation could be different, but usually there is CRS restart involved. Also there could be duplicate listener or listener_scan<x> seen on different cluster nodes.

Changes

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.