Are there specific benefits for a very large cluster to use ES 6.3.1 instead of ES 6.2.4?


We are currently working on creating an Elasticsearch deployment of around 350 Data Nodes, running on 70 physical servers (each server hosting 5 Nodes).

The cluster isn't in production yet, the app is still in development.
Currently we are developing based on Elasticsearch 6.2.4, but we have the option of using ES 6.3.1 instead (not any version more recent because of inner technical difficulties).

Since switching to 6.3.1 might cost us significant time in changing code and testing, I would like to ask:

Is there any known major issue in 6.2.4 that is solved in 6.3.1? Especially in regard to very large clusters.
I have read through the issues on Github and was unable to figure out whether the issues solved are crucial / relevant to our project or not.

Thank you

Why is each server hosting 5 nodes ?
You should look into the release-noted. If they are ok, you're ok.

Thank you for your reply. Let me clarify my question:

Is there a specific bug in 6.2.4 which is important to large clusters and is already solved in 6.3.1?

Any kind of memory leak bug, lesser performance of large indices, anything of that nature?

Regarding your question: each machine is hosting 5 nodes because these are large machines with 0.5TB of RAM.

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