From the the slow logs I see that index.search.slowlog.fetch and QUERY_AND_FETCH slow for queries. my understanding is this where the ES fetches the documents after it determines the ids in the QUERY phase.
I am thinking of keeping just ids in elasticsearch and move the document to redis , can i change elasticsearch to do the fetches from redis? or should I get back the ids from ES and then do a redis fetch?
From the slow logs fetches seem slow, I am seeing this might be to do with disk IO so instead of storing the documents in disk in ES thought it will be better idea to use ES for searching /faceting/ filtering but the actual document fetch could be from in memory datastore
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.