List of videos

This is the second addition of performance tricks one has never seen before. Let us take a look at some of the hidden trickery a PostgreSQL consultant can use to make life better. Everything from indexes to storage to query tuning to other tricks. In this talk there will be wisdom which is passed on from "PostgreSQL druid to PostgreSQL druid" only.

Let’s talk about what it’s like for women in tech! Taking inspiration from the most popular movie of 2023, Barbie, we unravel the challenging journey of women navigating a male-dominated tech world. In this engaging discussion, we share poignant stories that highlight both subtle and overt instances of gender bias, offering valuable insights into the persistent hurdles that women encounter. By examining these narratives, we aim to foster awareness, encourage reflection, and inspire action to create a more inclusive and equitable IT landscape for everyone. Join us for an eye-opening exploration that paves the way for meaningful change in Ken’s mojo dojo casa house.

Upgrading your high available, high load PostgreSQL databases can be a challenge own its own. When having customers around the globe with tight SLAs, the requirement arises to execute these upgrades with minimal or even without any downtime. This talk walks through how we at GitLab evolved from multihour downtime windows to not causing any customer facing interference at all.

Join me on a captivating journey as I share my experience as a fresh contributor to the PostgreSQL community during Google Summer of Code 2023. In this talk, I'll dive into the heartwarming support that I received when embarking on my open-source adventure. Through my contributions to the project, 'pg_statviz', I will explore the process, challenges, and triumphs that marked my initiation into the world of PostgreSQL. The PostgreSQL community is renowned for its vibrant and welcoming nature, and I'll be your guide to showcasing just how true this reputation is. I'll be talking about the help I recieved from the developers and community members, and how this effort made my path of contribution and learning easier. Central to my presentation is the 'pg_statviz' tool, a minimalist tool I was happy to help make better during my GSoC project. I will provide a comprehensive overview of its functionalities, detailing how it offers database administrators and developers insights into PostgreSQL by storing data and statistics about PostgreSQL. Crafted with the purpose of capturing both cumulative and dynamic statistics within PostgreSQL, this toolset offers the capability to engage in insightful time series analysis. Its companion utility is adept at generating visually engaging representations, honing in on specific time intervals extracted from the stored statistical snapshots. This empowers users with the ability to monitor the performance of PostgreSQL over time, opening doors to potential optimization and troubleshooting requirements.

pg_failover_slots is a new extension released this year that facilitates maintaining logical replication and logical change capture during physical replication failover and switchover. This is, arguably, an essential component of being able to use logical replication reliably. In this presentation, I will show the ideas and design behind pg_failover_slots, which challenges it solves, and how it can be configured. I will also discuss how it can be integrated into other replication management systems such as Patroni.

We all know that Postgres is a popular database for many different applications. Developers love the fact that Postgres "just works" out of the box. On the other hand, the default parameter values aren't always right for production, and you might not have a DBA on hand to configure them for you. Fortunately, you don't need to learn about all 350+ Postgres parameters to get the most out of your database. We'll look at the most important settings, and give some rules of thumb for tuning them according to your use-case, to make sure your database performs at its best. Then you can leave Postgres to "just work", and you can focus your efforts on developing your application.

For a few years we (Andres Freund, Thomas Munro, Melanie Plageman, David Rowley) have been working towards using asynchronous IO (AIO) and direct IO in Postgres. In this talk I'll describe what using AIO in postgres looks like, what we already have gotten into postgres (e.g. speeding up concurrent COPY by 3x in 16), what we are working to get into postgres 17, and what we hope to achieve longer term. The goal of using AIO and DIO in postgres is to improve throughput, decrease latency, reduce jitter, reduce double buffering and more.

Starting with PostgreSQL 16 there is a new catalog view called "pg_stat_io". Traditionally PostgreSQL did not provide much details when it comes to I/O statistics and this will change from version 16 on. This is all about extends, reads and writes, and statistics related to the buffer cache. By attending this talk you'll get a better understanding what PostgreSQL is doing I/O wise and we'll also cover direct I/O, which currently is a developer feature which is supposed to evolve over the next releases.

Geolocation data is a hot topic. A lot of data includes an address, a street name, a zip code, and a country. But where exactly is this address? And what's around this address? To solve these questions, addresses need to be resolved to geocoordinates. The better and more precise the address data, the better the chance to find the correct coordinates. This talk shows how addresses can be resolved using OpenSteetMap data, from within the database, using Python functions. We also talk about the accuracy of the data by comparing the results of different address formats to known coordinates.