Indexation issue

Hello,

We have a performance issue on our cluster elastic that doesn't succeed ingesting in real time our 3000eps.

The size of our indexes is around 200GB.
We have 8 nodes, and each of our index have 4 shards and 1 replica.
Our heap size on each of our data nodes is 16GB.
We have 32 GB available but reading elastic docs we have seen that lucenes processes need 50% of the RAM.

Does the number of shards impact the indexation process ?
We know that more shards you have more the research process is fast but what about the indexation ?
If we add shards, does the indexation will be faster ?

Thank you for your response
Pierre

Yes. The more primaries you have, the more parallel indexing the cluster can support.
But are you sure that the contention comes from elasticsearch?

How do you index your data?

May I suggest you look at the following resources about sizing:

https://www.elastic.co/elasticon/conf/2016/sf/quantitative-cluster-sizing

And https://www.elastic.co/webinars/using-rally-to-get-your-elasticsearch-cluster-size-right

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