PostgreSQL at 10TB and Beyond

In the discussion we will briefly mention the ordinary issues people run into when scaling to large volumes and workloads of data, and then we will focus on the problem solving process, using 4 problems and their solutions to illustrate. Two of the problems address a use case of PostgreSQL (job queueing) which has well known difficulties. The other two focus on advanced features. The problems discussed will be:

1. Autovacuum performance and planner limitations on large, high traffic, persistent queues.
2. Race conditions between snapshot creation and advisory lock release in a job queue.
3. Use of non-1NF designs and how to prevent data projection issues.
4. Development of new search criteria by extending SQL for semistructured data.

Along the way we will talk about the need for monitoring, testing, and even simulating database performance problems.

You can register for the event (free of charge) at the EventBrite page.

Posted by Edument AB.

Note: The PostgreSQL Global Development Group does not endorse any events run by third parties. No guarantee of the quality of events is offered whatsoever.