Time-series data is the backbone of many applications, tracking measurements from devices, financial transactions, and users at high frequencies. In nearly every application there is a need to frequently retrieve the most recent data point for each tracked item. As data grows, however, querying this information efficiently can become a bottleneck, even with properly tuned and indexed tables.

In this talk, I'll discuss the reason querying "last point" data in PostgreSQL can be problematic and demonstrate five approaches to keep your queries responsive. The size and scope of your data, as well as the development expertise on your team, will influence which approach is right for your situation. You'll leave the talk ready to evaluate your schema, indexes, and queries to keep your application running at peak performance.