Elasticsearch One Cluster or Multiple Clusters?


My colleagues and I wish to store time series data (such as logs), as well as user data in a Elasticsearch 7.2.0 cluster. We cannot decide if we should use one big cluster (to store both time series data and user data), or create two separate clusters (one for time series data, and one for user data). Is there any advantage to one approach or the other? Any thoughts, ideas, and things we should consider would greatly help.

Thank you,


Are the data somewhat related? If you don't want to search in both, and if the timeseries dataset could be huge, I'd probably create 2 clusters.
Could be then easier mentally to separate the concerns. Think about hot/warm/cold dedicated nodes...

If you are "short" on budget, then a single cluster.

At the end, I'd also look at cloud.elastic.co where you can easily build different clusters for various use cases.

My 2 cents.

Hi David,

Thank you very much for your help. Sounds like two clusters is the way to go!

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