I am migrating to Elastic Search 5.2.2 from 2.3.1, we are seeing performance Issue while indexing during Load Testing.
We do have 20000 to 40000 on an average id and its content being indexed.
Typically our Shard size is 25 G. We have 150 indexes
No of Shard=5
Replica = 1
No of Master=3
No of Data Node=3
Still not able to figure out performance reason so would like to understand following configuration idle for Production/Recommended in Prod
max number of files
vm.max_map_count
max user processes
Xms(jvm.options)
Xmx(jvm.options)
For Xms and Xmx, we usually recommend 1/2 of the machines ram, but no more than 31gb, so if you have 16gb of RAM, you would set these to 8gb (as an example).
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.