Master node missing

I have an ELK cluster, with two elasticsearch nodes, one with roles master, data and ingest, and the other with only data and ingest. There's also a script making backups each night to a shared disk on another server. I want to change the location for the backup to another shared disk, and ran into some issues. When I only changed the location in the script I got the following error message:

[backups] location [\\\\MYSERVER\\MyNewSharedDisk\\ELK_Cluster\\Backups\\snapshot_2021_05_28\\] doesn't match any of the locations specified by path.repo

So I figured I'd add the new location to elasticsearch.yml path.repo of both nodes. And after I do this I should do a rolling restart. But I noticed that the config for the master node doesn't have a configured. And according to this:

Master nodes must have a directory whose contents persist across restarts, just like data nodes, because this is where the cluster metadata is stored. The cluster metadata describes how to read the data stored on the data nodes, so if it is lost then the data stored on the data nodes cannot be read.

So now I'm worried that if I restart something will break. And I can't set without restarting either I think. What can I do to fix this, to change the backup location?

Elasticsearch version 6.6.2

(I'm new to elasticsearch, and recently took over this project, so I'm not entirely familiar with what I have running or with elasticseach terminology. The person who originally set it up is not available for questions)

Welcome to our community! :smiley:

You should upgrade, 6.6 reached EOL some time ago.

Just set path.repo and you will be fine.

There is a default that it will be using - should be /var/lib/elasticsearch on linux. That note is there partially because people were using non-persistent disk (eg with spot instances) and losing their clusters when the master node was terminated.

Ah ok great, thanks for the info!

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.