A brand new take on our Solr&Elasticsearch talk from last year! As the title suggests, this time we’ll dive deeper into how these two search engines scale and perform. And of course, we’ll take into account all the goodies that came with Elasticsearch and Solr since.
Both search engines are based on Lucene, so you’d expect similar numbers, but:
- at scale, small differences can give different numbers
- as with most functionality, Elasticsearch and Solr take different paths to achieve similar results, so you’d tune them differently
- their distributed models are quite different
We’ll show how you’d tune Elasticsearch and Solr for 2 common use-cases - logging and product search - and what numbers we got after tuning. Also, we’ll share some best practices for scaling out massive Elasticsearch and Solr clusters. For example, how to divide data into shards and indices/collections that account for growth, when to use routing and how to make sure that coordinating nodes don’t become unresponsive.
By the end you’ll see how Elasticsearch and Solr compare when you dive deeper into their functionality. You’ll know which important Lucene knobs to turn and how to do that in each search engine. Also, you’ll know how to use specific scaling features such as automatic rebalancing for Elasticsearch and shard splitting for Solr.