How (not) to Scale Elasticsearch for Data Analytics!
3rd December 2020
Speaker Time: IST: 16:00-16:30
Attendee Date: 3rd December 2020
Attendee Time: IST: 16:00-16:30 | SGT: 18:30-19:00 | AEDT: 21:30-22:00
Search is ubiquitous - From booking a cab on Ride sharing platforms to searching for a job on LinkedIn, Elasticsearch has emerged as a prominent solution for extensive data analytics and searches. Due it’s distributed nature and ability to offer powerful visualizations, it serves as a key tool for deriving actionable insights. And, the recent advancements in Elasticsearch using ML stands out due to its simplicity in data analytics.
While Elasticsearch’s out of the box defaults works for smaller implementations, the nuances of handling a massive deployment in order to scale for petabytes of data and billions of ingests & searches per day requires a deeper knowledge on Elasticsearch internals.
In this talk, we will cover:
- Learnings from our experiences in managing and scaling enterprise grade Elasticsearch clusters with specific details around the things that can go wrong and how we could mitigate them.
- Our inhouse security plugin that saved us 1 Million$ licensing cost.
- Gaining actionable insights using ML
- Demo of the features that we built on the top of Elasticsearch for scaling it.