Session
DevOps at Scale: Achieving a Zero-Downtime Migration for 80 Million Daily Requests
How do you pull off a seamless data platform migration in a mission-critical environment—without taking your systems offline? In this talk, we’ll delve into how our team leveraged DevOps and agile principles to execute a zero-downtime migration from AWS OpenSearch to Elastic.co’s Elasticsearch, handling over 80 million daily requests.
Attendees will learn how we used parallel ingestion pipelines, weighted routing for gradual traffic cutovers, and massive data synchronization to ensure a smooth transition. We’ll highlight the observability and feedback loops that guided our decision-making, along with the iterative practices that allowed us to pivot quickly and maintain confidence throughout the migration. You’ll walk away with practical insights on designing resilient architectures, mitigating risk, and applying DevOps best practices to tackle large-scale migrations—without leaving your users in the dark.

Kyle Hoehns
Staff Software Engineer and Teammate Pipeline Architect at Source Allies
Ankeny, Iowa, United States
Links
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top