Zero Downtime: Live Data and Server Migrations

Moving a live production dataset between servers is hard to do. I’m going to share a few examples of how I’ve been able to move data and workloads between servers without downtime, dropped writes, or stale reads. I had to break out more than a few Postgres tricks like logical replication, foreign data wrappers, and writable views, all of which we will be reviewing.

My goal is to share a few strategies that have worked and how they handle the common difficulties with moving live datasets. Please bring your own stories and questions and add to the conversation.