I've noticed when my ES (5.6) cluster is indexing lots of documents the number of queries being answered also increases. I index using the op_type "create" with predefined ids, so I guess this is just the cluster checking whether the document exists before indexing it, but I'm surprised this appear as queries and not get requests? Shouldn't get requests be more efficient since it's checking by id?
I use the Datadog metrics "elasticsearch.search.query.total.count" and "elasticsearch.search.query.time.count" to monitor the load.
If you are replacing a document with a static ID then it'll be directly routed to the required shard to do that. I don't recall if Elasticsearch does a direct get on the ID or if it uses a search.
That version is also EOL and no longer supported, you should be looking to upgrade as a matter of urgency as there's literally years of improvements you are missing out on.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.