Understanding a project’s ecosystem is the main barrier to entry. I’ve had community roles for different projects (Passenger app server, k6) and technology areas (Ruby, DevOps, Microsoft) but I found the Postgres ecosystem one of the most difficult to grasp.
However, the project and the people grew on me and I believe that in order for it to keep up with the growing demand as the most popular database according to StackOverflow, Postgres needs to onboard new contributors, and it needs to do so quickly.
There are plenty of people who want to contribute, but don’t know where to start. I think I do. Join me for a module I developed for inhouse training at EDB, that I’ll have open sourced by the time PGDay Austria takes place, in line with the project’s ethos.
PostgreSQL’s autovacuum tuning in the cloud era is a delicate balancing act. It demands careful juggling of application performance, cloud costs, and maintenance(autovacuum) efficiency. Databases on bare-metal enjoyed a happy flow with superior throughput and ultra-low latency. In contrast, the cloud era has suddenly introduced performance speedbreakers, even as it simplifies management for DBAs.
This talk will unravel how to optimize autovacuum settings for peak efficiency across AWS, GCP, and Azure. Discover strategies to overcome throughput and resource limits without busting your cloud budget, keeping your cloud FinOps team smiling all along.
A decade or two ago, VACUUM in PostgreSQL was a simple procedure that read an entire table and then removed no longer visible row versions from the table and its indexes. Every 200 million transactions, the table was traversed again to freeze old transaction numbers so the transaction counter wouldn’t overflow.
Over time, this seemingly simple procedure has been optimized. VACUUM now knows which pages to visit, both for vacuuming and freezing. Traversals can be accelerated by omitting indexes and TOAST tables. Various phases can utilize parallelization. Autovacuum now knows how to handle INSERT-only tables. Emergency freeze runs are far less intimidating than they used to be.
In this talk, we’ll explore these modern VACUUM features and show how DBAs can take advantage of the improved performance.
This talk explores a new way to scale read-heavy PostgreSQL workloads through SQL query caching and incremental view maintenance (IVM). Attendees will learn how our dataflow-based engine enables low-latency reads using partial materialization, without modifying application code. Real-world examples and architectural deep-dives illustrate how teams can achieve significant performance gains with minimal operational overhead.
Open source… What does that actually mean in real life? How do you turn a buzzword into something real that creates value and solves real problems? What does open source and the PostgreSQL database world mean for everyday life?
This session explores how to integrate PostgreSQL and Flyway into CI/CD pipelines to enable safe, automated database deployments. Learn best practices for managing schema migrations, version-controlling your database, and synchronizing database changes with application releases. A practical guide to making database delivery a seamless part of your DevOps workflow.
