Elasticsearch write performance

helllo,when i write data to Elasticsearch using the bulk api,which process is flink consume the data of kafka then write to Elasticsearch.sometimes the write performance has Large fluctuations,I want to kown when the elasticearch master node to publish the clusterstate to other nodes which maybe block the write performance? I konw when the no master block param value is write will block the write. I want to have a help about the question how to improve the writing performance. and the picture is the log from the elasticearch master node.

What is the full output of the cluster stats API? What is the size and structure of the cluster? What type of hardware and storage is used?

thanks ,I just want to kown the case,then the master publish the clusterstate data will block or reduce the writing rate?

If you are creating a new index to write to it has to be created and made available before it can be written to, which requires the cluster state to be updated. Any requests writing to the new index would therefore be delayed, which could affect the write rate.

Thanks for your answer

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