Cluster failed to come online."8.18.5
The casa.db.script configuration file acts as the source of truth for node roles and contains incorrect entries. Specifically, the primary node is erroneously configured with "is_admin_node": false. As a result, the cluster manager fails to recognize an admin node during startup, preventing the cluster from coming online.
Location of the casa.db.script file.
# /storage/db/casa/webapp/hsqldb/Note: This is a simple script for easier readability:
# sed -nre "/clusterMembership/ s/^[^']+'([^']+)','([^']+)'.*/\2/p" /storage/db/casa/webapp/hsqldb/casa.db.script | python -m json.tool
It extracts the second field from lines containing clusterMembership in the casa.db.script file and formats the output as pretty-printed JSON.
To fix the issue the files need to be updated correct entries:
Note: Each of the below steps must be performed on every node.
root using SSH.# service vmware-casa stop# cd /storage/db/casa/webapp/hsqldb# cp -p casa.db.script casa.db.script.bkpcasa.db.script file using a text editor (using vi)."is_admin_node": false,"ip_address":"<REDACTED_IPS>""is_admin_node":true# service vmware-casa start