Identity Manager node has high replication delay but shows as up

Products

• VCF Operations/Automation (formerly VMware Aria Suite)

Issue/Introduction

With the poolnodes command, one of the standby nodes shows as UP, but also has high replication delay, e.g.:

• Inter-node password-less SSH works fine among all nodes.
• All nodes' /db/data directory and sub-files have the correct postgres:users ownership.

 

Environment

Identity Manager 3.3.7

Cause

PostgreSQL database replication de-synchronization on the affected vIDM node.

Resolution

Manually trigger a database re-sync for the affected node using the pcp_recovery_node utility.

1. Obtain the specific node ID of the delayed node from the poolnodes output.

2. Execute the following command to force the re-sync, replacing <FAILED_NODE_ID> with the actual ID:

/usr/local/bin/pcp_recovery_node -w -h localhost -U pgpool -p 9898 -n 0

Confirm again with the poolnodes command that the replication delay is gone, e.g.: