This presentation was given by Alexander Reelsen at the Munich NoSQL Meetup on March 27, 2014.
This presentation starts with a quick introduction to Elasticsearch. We then dive into a full view of Elasticsearch: hardware, operating system, JVM and garbage collection down to its libraries, all from a performance point of view. This talk explains better known and less well known things you should be aware of when maintaining performance in a distributed system. While these topics are explored through the lens of Elasticsearch, the tips and tricks are valuable to anyone interested in performance for distributed systems, whether or not they are using Elasticsearch.