talk-data.com

Please join us for a night of PostgreSQL!

PostgreSQL is awesome - and out of the box performance could certainly be customized! In this talk\, we’ll dig into the key fundamentals for PostgreSQL performance.

We’ll approach this from two angles - first memory cache and data management. There are a couple tricks for getting the most out of PostgreSQL memory without throwing more RAM (and money) at the problem. Second, we’ll look at PostgreSQL’s internal EXPLAIN tools to diagnose issues with queries, test improvements, and make your queries faster. You’ll leave this talk with a better grasp on what helps PostgreSQL perform and how to get started on your own performance improvements.

RSVP details: This meetup will be held at the Snowflake Manhattan Office at 114 West 40th Street, NY, NY 10018.

Building security will require a complete list of RSVPs more than 24 hours ahead of time. Space is strictly limited at 30 guests so please change your RSVP if you cannot make it. Guests will be registered in the system via email and must fill out a check in form. At the time of the event, guests will check in and present identification to lobby security, and then be let up to the meeting suite.

Please specify your dietary needs in the RSVP form as well so we can plan the food accordingly.

Details: Tuesday, September 30th. 6-7:30pm Snowflake Manhattan Office at 114 West 40th Street Snacks and drinks will be served

Sponsor: Snowflake

Additional Info: This is our first meeting after a very long break and we are so excited to meet and kickstart our group again. If you have ideas for future group spaces, events, or talks - please reach out. This group is open to anyone interested in PostgreSQL and we welcome speakers and attendees at all points in their data careers.

We will be hosting our September Meetup.

Thanks to Amazon for sponsoring us. Please note that you have to show sign up with your real name and show a Photo ID to the building security before entering the building. This is a requirement from the building security. If you are not ok with that, please do not sign up. You can not bring a +1, everyone needs to sign up. Please only sign up if you plan to come, we will keep notes.

Speakers: Divyendu Singh: How LLMs Help Us Manage PostgreSQL at Scale Marko Coha: Postgres in the Cloud

Additional information for Meetup attendees are available here: https://andreas.scherbaum.la/post/2025-05-05_attending-organizing-or-speaking-at-a-meetup/

How LLMs Help Us Manage PostgreSQL at Scale: The Xata Agent is an open-source AI PostgreSQL expert. It monitors Postgres clusters by analyzing metrics and logs, checking pg_stat_statements for slow queries, running EXPLAIN and interpreting the output. We have been using the Agent internally on a large fleet of production PostgreSQL clusters, and we have gathered valuable insights, including: - Comparing different models and providers and their benefits. - Combining cloud hosting observability data with Postgres tools. - The cost implications of using LLMs for database monitoring. - Testing the agent’s effectiveness in detecting performance regressions.

Postgres in the Cloud: This presentation provides a comprehensive overview of hosting PostgreSQL in the cloud, examining both strategic benefits and potential trade-offs. Attendees will first gain an understanding of the key reasons organizations choose cloud-based PostgreSQL deployments, including scalability, operational efficiency, and cost optimization. We will explore how cloud platforms can simplify database administration, accelerate time-to-market, and support dynamic workloads, while also enabling global accessibility and high availability.

We are happy to announce our September PostgreSQL Meetup!

AWS gracefully offered to host us on September 18th, starting 18:00.

We have two great talks this time!

Exploring index infrastructure: Implementation of an opclass for UUID v1 support in PostgreSQL

by Nick Babadzhanian.

We'll touch on a number of topics ranging from WAL amplification and performance impact of data distribution to extension development and of course opclasses and their role in the index infrastructure of PostgreSQL.

We will touch on the internals but keep it friendly for a wide audience.

and

Patroni 3.0: What's New and Future Plans

by Polina Bungina and Alexander Kukushkin.

Patroni is one of the most popular and advanced solutions for PostgreSQL high-availability. It is integrated with a variety of distributed configuration stores allowing Patroni to work with PostgreSQL cluster information in a consistent way ensuring that there is only one leader at a time. Unlike the majority of existing solutions for automatic failover, Patroni requires a minimal effort to configure the HA cluster and supports auto-discovery of new nodes.

This talk will consist of a short introduction to the ideas behind Patroni, overview of the major features that we have released lately, cover some interesting failure scenarios which were fixed, and share some plans for the future development. Of course, the talk will not be complete without showing a live demo of some brand-new features.

Location: Amazon c/o Rocket Tower (BER16)

Charlottenstraße 4

10969 Berlin

Please note that you need to provide a registration and a photo ID for entering the building. Check-in will need approx. 30 minutes, please plan accordingly.

Food and soft drinks will be provided, alcohol is not permitted during the event.

Amazon is running a 1-day training event on 19th September. If you would like to participate, please contact your AWS SA/TAM/AM for more details.