After restarting the cluster, Cassandra will not start on nodes other than the primary node

Products

• VCF Operations/Automation (formerly VMware Aria Suite)

Issue/Introduction

After restarting your Aria Operations for Logs cluster, Cassandra will only start on the primary node leaving the web UI inaccessible.  

• Running the command nodetool-no-pass status only shows the primary node being up (UN status in the first column of the output)
• Following "FAILED: Unable to get user data. Possible Cassandra is down" - Aria Operations for Logs does not resolve the problem
• Updating the certificate with SSL certificates are expired for Aria Operations for Logs (Formerly Log Insight) does not resolve the problem
• /storage/core/loginsight/var/cassandra.log on the primary node shows errors similar to:
  
  ERROR [main] 2025-09-11T13:04:49,477 CassandraDaemon.java:900 - Port already in use: 7199; nested exception is:
        java.net.BindException: Address already in use (Bind failed)
java.net.BindException: Address already in use (Bind failed)
        at java.net.PlainSocketImpl.socketBind(Native Method) ~[?:?]
        at java.net.AbstractPlainSocketImpl.bind(Unknown Source) ~[?:?]
        at java.net.ServerSocket.bind(Unknown Source) ~[?:?]
        at java.net.ServerSocket.<init>(Unknown Source) ~[?:?]
        at javax.net.DefaultServerSocketFactory.createServerSocket(Unknown Source) ~[?:?]

Environment

Aria Operations for Logs 8.18.x

Cause

An orphaned Cassandra process is using port 7199 and cannot be stopped by systemctl stop loginsight or /usr/lib/loginsight/application/sbin/li-cassandra.sh --stopnow --force

Resolution

To find and stop the orphaned Cassandra process to allow the proper Cassandra process to start:

1. Log into the primary Aria Operations for Logs node and run systemctl stop loginsight
2. Run /usr/lib/loginsight/application/sbin/li-cassandra.sh --stopnow --force
3. Run ps aux | grep java
4. Note the process ID (PID) of any java process still running. It will be the number after root in the output, for example:
   root      3312
5. Run kill -9 <PID> for all running java processes. (There is likely only one). 
6. Run systemctl start loginsight
7. Repeat steps 1-6 on the other nodes in sequential order, then run nodetool-no-pass status afterwards to ensure prior nodes now show the UN status before proceeding to the next node.