We are moving away from our legacy version of elasticsearch 2.0 (with marvel and alerting) to 5.0 with xpack.
I was wondering what is an easy and good way to performance test elastic? I want to see how it affects the performance from upgrading, and to see if the xpack plugin requires us to have a better servers because of the additional plugins.
Our server configuration as of now:
3x machines (1 node per machine) with 4 cores and 32GB of ram
I am looking to change that to:
each machine have 3 nodes, using containers. each node will have 2 cores and 9GB of RAM. What do you think?
I do not see the point in having a larger number of smaller nodes on the same hardware as it increases overhead and complexity. The point where you start having more than 1 node per server is generally when you have more than 64GB RAM on the host and want to maximise heap space, which does not seem to be the case here.
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.