Moving Graylog Master Server

I have a graylog cluster setup in the following config:
2x MongoDB VMs
2x Graylog physical systems - one is master
6x Elasticsearch physical systems
Things are running smoothly, however as we’re increasing load on the Graylog boxes we want to ensure that the master services, and GUI are out of the path of log processing. The end goal is two systems with physical resources dedicated to Graylog processing logs. One VM will be dedicated to hosting the web console and performing the tasks that the master system performs.

The question I have is, what’s the best way to transfer the role? Should I simply join the VM to the cluster as a secondary, and then when I’m ready to make the shift:

  1. Mark the VM as master in the Graylog config.
  2. Mark the current master not master in the Graylog config.
  3. Restart Graylog on the former master.
  4. Restart Graylog on the VM master

Will that cause an issue when the VM tries to assume the master role? I’m assuming all of the masters tasks are saved in the MongoDB so the new VM will simply pick up where the former master left off.

your choosen way would be the best to follow.

1 Like

Thanks Jan. I’ll give it a shot and report back if I have any issues for other folks.

Just wanted to ping back. I brought the VM online as a cluster member and ensured that the system worked as expected by navigating to its FQDN and then hitting the GUI as much as possible. Once I was sure that the system was working as expected I adjusted the server.conf file as outlined above and restarted Graylog service on the master first and then the VM master second. All told it took about 30s and the new VM was listed as master and there was very little impact to message processing.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.