Technical Tip: Node.js-charting down after successful upgrade due to hostname mismatch

Description

This article describes how to troubleshoot and recover the Node.js-charting process after a FortiSIEM upgrade when the upgrade completes successfully but the Node.js-charting service remains down.

Scope

 FortiSIEM v7.x.

Solution

During a FortiSIEM upgrade, the Node.js-charting process may fail to start even though the upgrade completed successfully.

The following Fortinet article was initially used for troubleshooting: Troubleshooting Tip: How to troubleshoot Node.js-charting process. As part of the troubleshooting process, the following commands were executed:

cd /opt/charting
./redishb.sh | tee /tmp/rhb.out

During execution, the following error was observed:

Error: getaddrinfo ENOTFOUND FSMSUPERVISOR
    at GetAddrInfoReqWrap.onlookupall [as oncomplete] (node:dns:122:26)

Emitted 'error' event on Commander instance at:
    at RedisSocket.<anonymous> (/opt/charting/node_modules/@redis/client/dist/lib/client/index.js:422:14)
    at RedisSocket.emit (node:events:519:28)
    at RedisSocket.#onSocketError (node:internal/modules/@redis/client/dist/lib/client/socket.js:166:18)
    at process.processTicksAndRejections (node:internal/process/task_queues:103:5)
    at async Commander.connect (/opt/charting/node_modules/@redis/client/dist/lib/client/index.js:185:9)
    at async file:///opt/charting/SVGCluster.js:31:1 {
  errno: -3008,
  code: 'ENOTFOUND',
  syscall: 'getaddrinfo',
  hostname: 'FSMSUPERVISOR'

The error indicated a hostname resolution failure from the Node.js Redis client used by SVGCluster.js. To verify the process state, run:

ps -ef | grep SVGCluster

The output confirmed that no active SVGCluster.js process was running. Further investigation of the /etc/hosts file identified a hostname mismatch.

Existing entry:

x.x.x.x FSM_SUPERVISOR

However, the application was attempting to resolve:

FSMSUPERVISOR

The root cause was identified as a hostname mismatch between the configured hostname and the hostname expected by the Node.js charting services.

To resolve the issue, add the missing hostname entry to /etc/hosts.

Note: Replace x.x.x.x with the current Supervisor IP address:

echo "x.x.x.x FSMSUPERVISOR" >> /etc/hosts

After updating the hosts file, rerun the following:

cd /opt/charting
./redishb.sh | tee /tmp/rhb.out

The Node.js-charting services should start successfully afterwards.

To verify the process is running properly:

ps -ef | grep SVGCluster

Expected result:

• Multiple SVGCluster.js processes should appear.

• phstatus should show Node.js-charting as running.

Summary:

• Upgrade completed successfully but Node.js-charting remained down.

• Troubleshooting identified a hostname resolution failure:

Error: getaddrinfo ENOTFOUND FSMSUPERVISOR

• The root cause was a mismatch in /etc/hosts:

Existing hostname: FSM_SUPERVISOR
Expected hostname: FSMSUPERVISOR

• Adding the missing hostname entry resolved the issue and restored the Node.js-charting services successfully.