FYI we’ve renamed ELK to the Elastic Stack, otherwise Beats and APM feel left out!
You don't need that much for masters, coordinating or ingest. Generally aim for;
Masters - 4-6GB heap
Coordinating - 8-16GB heap
Ingest - 8-16GB heap
For each of these, assuming they are dedicated nodes types, you can exceed the standard 50% heap:off-heap rule as they don't need to worry about caching the underlying file structures. In these instances you can use 75% of total memory for heap.
@warkolm I apologize, perhaps my message was a bit misleading... let me try again)
Elasticsearch (master and data) node runs respectively on it's own dedicated VM w/ ~54GB of memory, I allocate 32766mb for each node and now I'm trying to comprehend how much memory "coordinating only" node would requires especially considering Kibana would run on same VM as well, same goes for ingest node as well; memory requirements (obviously no Kibana).
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.