Changes to node configuration can temporarily cause a duplicate node entry on the clustering page

Summary

Changes to a node in a Crowd Data Center cluster can cause a duplicate entry to be seen in the clustering admin page <crowd-base-url>crowd/plugins/servlet/cluster-monitoring. This is particularly confusing because it appears as if 3 nodes were in the cluster when really there are only 2. (see screenshot)

Environment

• Crowd
• Data Center

Steps to Reproduce

1. Create node 1 in a Crowd Data center environment with DC License via these steps https://confluence.atlassian.com/crowd/installing-crowd-data-center-935369773.html
2. Configure and bring up node 2
3. Visit the cluster page and note that two nodes appear in the cluster: <crowd-base-url>crowd/plugins/servlet/cluster-monitoring
4. Visit the node 1 install directory <crowd-install>/apache-tomcat/bin and edit setenv.sh file
5. Note the following values JAVA_OPTS="-Xms128m -Xmx512m. Edit the file and change* -Xms128m* to -Xms512m
6. Save the file and restart Crowd on Node 1
7. When the node is online, visit the clustering page: <crowd-base-url>crowd/plugins/servlet/cluster-monitoring

Expected Results

You will see both nodes in the cluster

Actual Results

A duplicate entry appears for Node 1 and it appears as if three nodes are in the cluster instead of two.

Workaround

The duplicate entry will eventually disappear from the clustering page.

Notes

The screenshots are a before and after snapshot of the cluster page