Wayfair has been using the Elastic Stack as a centralized logging platform since 2014. Our infrastructure has over 2,700 instances with 500 TB SSD storage and supports over 3 billion logs a day. In this talk, we will go over how we migrated the on-prem platform to the cloud.
We will cover:
* How Wayfair is using terraform, consul and puppet to automate the deployment and configuration of Elasticsearch Clusters in GCP * How we designed the clusters to be cost-effective. * Will go over shard awareness, index settings, instance types, and storage types * How we leverage local SSDs for high throughput and yet overcome the trade-offs in availability, durability, and flexibility * How we are using hot and cold architecture to improve our query performance
Elastic's Solution Architect team will also cover "What's new with Elastic?" We will provide updates to our most recent releases and a look into our roadmap.
About the speakers:
Zubin Rupani is a DevOps engineer at Wayfair. He has been a part of the Wayfair Infrastructure Engineering team for over 3 years. He is leading the project to migrate Wayfair's Elastic Stack to GCP. He is lives in greater Boston with his fiancée and dog. In his free time, he loves to read about trending distributed systems, playing video games, watching soccer and hiking.