I just have some questions in regards with the different roles of Elasticsearch (v5.6). We currently separate all the roles of the servers to Ingest, Data and Master. I have some confusion in regards with how Ingest node behaves.
We're using the Pipeline feature during ingestion, but we also set our Java clients to search and retrieve data with the Ingest node as well. During the search process, does the Ingest node acts like a coordinating node which is responsible for scatter phase?
does the Ingest node acts like a coordinating node which is responsible for scatter phase?
All nodes are coordinating nodes in that every node is capable of executing a search and collecting the results.
"Coordinating only nodes" are so named because they don't perform any other tasks other than coordinating.
It is reasonable to have your clients connecting to ingest-capable nodes. It's also reasonable to separate them, depending on your node count and work load.
Awesome! It looks like my assumption was correct. The doc is a bit misleading about the having specific tasks for each roles like data node is responsible for search.
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.