Elastic Silicon Valley User Group: https://www.meetup.com/Silicon-Valley-Elastic-Fantastics/events/233535310/
Benchmarking Elasticsearch for your use case with Rally.
How many shards should my index have? How large should my shards be? How many shards should I have per node? These and many other questions about sizing an Elasticsearch cluster have the common answer of “it depends”. So how do you go about finding the answer for your use case? In this talk we’ll look at how we answered these question at Tubular Labs.
Matt Delaney is a production engineer at Tubular Labs. He enjoys the challenges of working with distributed systems and has worked with Elasticsearch for 3 years (two primarily for a Logging use case and one primarily for search).
Using Elastic to search over 2.5 billion videos
At Tubular Labs, we use Elasticsearch to drive several core products used by many online video industry leaders across the globe. This talk covers techniques used in Tubular to index hundreds of millions documents daily and searching through terabytes of data in real time. In addition, we’ll go through several cases on how we use ElasticSearch to simplify our architecture and reduce hardware costs.
Stas Starcha is an engineering manager at Tubular Labs working with a team of talented engineers who design and build real-time services using distributed databases. He is mainly inspired by three things: finding 10x improvements, music production, and snowboarding.
If you want to learn more visit us at https://www.elastic.co/learn
You can watch this video also at the source.