FireflyPGConf.DE 20242024-04-122024-04-121https://www.postgresql.eu/events/pgconfde2024/schedule/09:0000:10Ballsaal 1Openinghttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5509/Andreas Scherbaum09:1000:45Ballsaal 1PostgreSQL-As-A-Service: Vergleich von Cloud ProvidernDer gemanagte Betrieb in der Cloud von Datenbanken im Allgemeinen und von Postgres im Speziellen wird immer üblicher. Die drei Hyperscaler bieten mit Amazon RDS, Google Cloud SQL und Microsoft Azure Postgres jeweils einen Managed Postgres Service an.
Alle drei bieten Umgebungen, die via Web-Interface oder API Konfigurations-Änderungen und sonstige Einstellungen erlauben. Sie erledigen auch Backups (und optional Hochverfügbarkeit), sowie sonstige automatische Wartungen wie das Einspielen von Patch-Releases. Schließlich beiten sie eine Vielzahl von Postgres-Erweiterungen an. Auf der anderen Seite bieten sie keinen Superuser-Zugang zu ihren Postgres-Instanzen, was andere Dinge schwieriger macht.
Dieser Vortrag wird die Gemeinsamkeiten und Unterschiede zwischen den drei Managed Postgres Anbietern vorstellen. Es wird außerdem ein quantitativer Einblick in verschiedene Punkte gegeben wie Community-Engagement, das Verhalten bei Speichermangel, sowie den Patching-Zyklus bzw. wie lange es dauert bis neue Major-Versionen unterstützt werden.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5418/Michael Banck09:1000:45Ballsaal 2Elephant in a nutshell - Navigating the Postgres community 101If you are a newcomer to PostgreSQL or want to be more active within the community you need to understand how the community works, what different roles are, where to look for help and who stands behind the “PostgreSQL community”.
This talk will provide an overview of:
- different roles within the community,
- communication channels where one can look for help, conferences and events;
- user groups - How to start a user group? What are the steps for having a successful meetups?
- What are some challenges and where to get support?;
- postgres development - commitfest, roles and responsibilities and the overall process.
Presentation will have references to other talks, guides and materials that dive deeper into specific processes within the community.
The presentation can be served as a toolbox for anyone who is looking to join and contribute to further development of The World's Most Advanced Open Source Relational Database.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5349/Valeria Kaplan09:1000:45Ballsaal 3Postgres native incremental backups and the future of backup toolsRobert Haas's commit which adds native incremental backup in Postgres has brought up a lot of interest since last December. But it’s not a one-fits-all solution. Rather, it's a very flexible feature that brings up a lot of opportunities for building solutions that adapt to different needs.
I’ll introduce upcoming incremental backup in Postgres, both how to backup with pg_basebackup, and how the new pg_combinebackup tool works for backups and restores. We will go into possible architectural solutions that can be taken once the feature is released, and how the backup tools in the Postgres ecosystem can benefit from it.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5445/Martín Marqués09:5500:30Ballsaal 1Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5506/09:5500:30Ballsaal 2Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5507/09:5500:30Ballsaal 3Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5508/10:2500:45Ballsaal 1Geodaten-Management mit PostGISIn der Welt der Geographischen Informationssysteme (GIS) wird die PostgreSQL-Datenbank in Kombination mit PostGIS häufig zur Speicherung räumlicher Daten verwendet. Die PostGIS-Erweiterung, die in den letzten 23 Jahren von einer großen wachsenden Gemeinschaft gepflegt und erweitert wurde, macht PostgreSQL zu einer räumlichen Datenbank. Im Wesentlichen ermöglicht PostGIS räumliche Daten in PostgreSQL zu speichern und zu nutzen.
Dieser Vortrag beschäftigt sich mit dem zentralen Teil, für den PostGIS verwendet wird: die Geodaten. Wir werden unterschiedliche Möglichkeiten aufzeigen, um räumliche Daten in einer performanten Art und Weise mit PostGIS zu speichern. Eine wichtige Rolle spielt hierbei sowohl die effiziente Organisation über das Datenmodell, als auch die Verfügbarkeit für Datenexporte und Datenabfragen von häufig verwendeten GIS-Clients (z.B. QGIS oder MapServer).
Anhand eines großen laufenden Projektes des Schweizer Bundesamtes für Landestopografie (Swisstopo) werden wir verschiedene gängige Datenmodelle illustrieren und deren Vor- und Nachteile diskutieren. Einige dieser Beispiele bestehen aus sehr komplexen Datensätzen mit einer großen Anzahl an geographischen Objekten. Die Art, wie die Daten gespeichert werden, ist insbesondere für eine performante Visualisierung entscheidend, was wir an dem Beispiel der nationalen Geodateninfrastruktur map.geo.admin.ch zeigen werden.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5470/Dr. Andreas JobstMarion Baumgartner10:2500:45Ballsaal 2Calculating the future: how to model PostgreSQL performanceA long time ago in a galaxy far, far away, a group of adventures went into a
deep dark cave on a distant planet to find a famous Prophet and ask him
Important Questions. The Prophet listened, took some time, then answered that
their database is going to do 3k IOPS.
While still mostly in the area of science fiction, the idea of predicting how
the database is going to behave under certain conditions is very attractive. In
this talk we discuss possible approaches for modelling, taking a simplified
PostgreSQL index under a load as an example -- and try to simulate and verify
the resulting IO profile. This journey will bring us to many places, going as
high as building a mathematical model, and as low as figuring out how BTree
page split is working.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5300/Dmitry Dolgov10:2500:45Ballsaal 3Catch Me If You Can - Change Data Capture with PostgreSQL and DebeziumLegacy database infrastructure often are batch driven, which means that DBAs often need to administer and control huge amounts of scripts, timely dependent on each other. This creates complicated chains of data processes, delays which huge datasets and probably not easy to handle database load and operational caveats. This doesn't scale well if datasets keep growing and more specialized data silos are added.
Modern infrastructure tries to decouple these dependencies and integrate data processes via streaming technologies. PostgreSQL and Debezium allows to create such architecture on top of complete opensource components. This talk gives an overview on how these techniques can be integrated in a stack with PostgreSQL, discusses the architecture and possible caveats which can arise.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5475/Bernd Helmle10:2500:45Sponsor TrackPostgreSQL Tooling for the CommunityAt CYBERTEC, we've carefully curated a wide range of Open Source tools designed to simplify PostgreSQL operations for users. Whether you need to monitor PostgreSQL performance, manage permissions, or tackle any other task, we're here to provide the ideal Open Source solution. Join us for a journey through our Git repositories, unveiling the innovation behind PostgreSQL's efficiency. Welcome to a preview of what the future holds for PostgreSQL management.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5517/Hans-Jürgen Schönig11:2000:45Ballsaal 1Sicherheitsattacken auf PostgreSQLIn diesem Vortrag betrachte ich verschiedene Methoden, um unter Ausnützung unsicherer Konfiguration oder unsicherer Objektdefinitionen Schaden zu verursachen:
- Verfügbarkeit beeinträchtigen ("denial of service")
- Zugang zur Datenbank erschleichen
- Zugriff auf verbotene Daten erschleichen
- Superuser werden
Wer die Gefahren kennt, kann sich davor schützen!https://www.postgresql.eu/events/pgconfde2024/schedule/session/5232/Laurenz Albe11:2000:45Ballsaal 2Bringing Vectors to PostgresPostgres does not yet have native vector capabilities (as of Postgres 16) and pgvector is designed to fill this gap. You can store your vector data alongside the rest of your data in Postgres and do vector similarity search while still utilizing all the great features Postgres provides.
The pgvector extension integrates seamlessly with Postgres – allowing users to leverage its capabilities within their existing database infrastructure. This simplifies the deployment and management of AI applications, as there's no need for separate data stores or complex data transfer processes.
In this talk, we will learn how to generate and store vector embeddings in Postgres. We will discuss indexing (IVFFlat, HNSW) the embedding data and illustrate how to run a similarity query on our embeddings.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5291/Gülçin Yıldırım Jelinek11:2000:45Ballsaal 3pg_upgrade like a boss!Upgrading a database cluster can be a complex and time-consuming process. While there are several methods for upgrade, such as dump/restore and logical replication, pg_upgrade is a powerful tool that allows users to upgrade a cluster with acceptable and predictable downtime.
In this talk, I will explore the benefits of using pg_upgrade for major upgrades, particularly in large, high-availability (HA) clusters with multiple nodes. I will provide a list of pre- and post-upgrade checks and tasks to ensure a smooth and successful upgrade process, and describe common pitfalls and how to avoid them. Through real-world examples and case studies, I will demonstrate the scalability and efficiency of pg_upgrade+rsync approach for major upgrades, and provide tips and best practices for using those tools in various environments.
Last, but not least, I want to raise one the most important and probably at the same time undervalued issue - lost statistics and how to deal with it.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5484/Alexander Kukushkin11:2000:45Sponsor TrackPostgreSQL in the Banking and Public Sector in GermanyA short introduction of the actual database market situation in Germany and how PostgreSQL is impacting the market with examples of the banking and public sector.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5520/Diego Calvo de Nó12:0500:45Ballsaal 1Lunchhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5510/12:0500:45Ballsaal 2Lunchhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5511/12:0500:45Ballsaal 3Lunchhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5512/12:5000:45Ballsaal 1KI ganz praktisch - Nutzung von der KI hinter ChatGPT mit aktuellen Daten aus PostgreSQL-DatenbankenIn diesem Vortrag beleuchten wir, wie die KI-Technologie, die hinter ChatGPT steht in Verbindung mit Echtzeitdaten aus PostgreSQL-Datenbanken genutzt werden kann, um maßgeschneiderte und dynamische Anwendungen zu entwickeln. Wir demonstrieren, wie die OpenAI-Schnittstelle effektiv in Softwareprojekte integriert wird und dabei direkt auf aktuelle Daten aus PostgreSQL-Datenbanken zugreift, um adaptive und intelligente Systeme zu schaffen.
Die Teilnehmenden erhalten eine Einführung in die KI hinter ChatGPT und lernen die OpenAI-Schnittstelle kennen. Besonderes Augenmerk liegt auf der praktischen Anwendung: Wir zeigen, wie man KI-Modelle mit Live-Daten aus PostgreSQL-Datenbanken verknüpft, um Echtzeitanalysen, personalisierte Benutzererfahrungen und automatisierte Prozesse zu ermöglichen.
Ziel des Vortrags ist es, den Teilnehmenden konkrete Methoden an die Hand zu geben, mit denen sie KI in ihre eigenen Projekte einbinden können, insbesondere durch die Nutzung von aktuellen Datenbankinformationen, um die Relevanz und Effizienz der KI-gestützten Anwendungen zu maximieren.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5362/Wolf G. Beckman12:5000:45Ballsaal 2StandBy-Database Shootout - PostgreSQL vs. OracleUsing PostgreSQL in high-available environments is always based
on replication and standby databases. There isn't and probably won't
ever be an equivalent alternative to solutions like Oracle Real
Application Cluster regarding scaling out the database service on
multiple active nodes. Multi-master-replication solutions with PostgreSQL
are somehow pointing in this direction but only in special scenarios.
As the most hyped target for migrations from Oracle the comparison
to Oracle Data Guard is inevitable.
In this talk a direct comparison will show similarities, differences
and the advantages or disadvantages of both worlds in detail including
the relevance of additional PostgreSQL solutions like Patroni or repmgr.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5400/Dirk Krautschick12:5000:45Ballsaal 3WORK_MEM the good, the bad and the uglyPostgres WORK_MEM is an important parameter to control the private memory allocated to the worker process but is not that often discussed in-depth at pgconf / pgday events.
Over the last decade, I have worked on migrations of literally 1000s database clusters (in data centres around the world) from Oracle to Postgres. I'm now working exclusively with Postgres on some exceptionally high workloads.
I have found tuning WORK_MEM generally quite intuitive, but there are some edge cases, and occasionally, things can get very ugly (at least in the DBaaS world).
I will demo the normal tuning processes I use and then go onto some interesting edge cases I have seen.
Finally, I want to cover some defensive programming best practices to avoid ugly edge cases.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5409/Dave Pitts12:5000:45Sponsor TrackOperational AI Workload Means Postgres Database WorkloadThe massive interest in AI solutions recently sparked a huge wave of AI projects. We now enter a second phase where valid AI projects look for operational enterprise environments, as shown by the big hype for vector databases and feature stores. This phase is the hour of databases, the recognised bedrock for enterprise data management environments, where Postgres is a front runner. AI workloads are fully tied to data: they start with data, run on data and produce data. We will review popular AI application flows and their strong ties to data, and then show how Postgres' strengths perfectly fit mission critical enterprise AI solutions. Finally we will share our plans for further evolution of Postgres towards a fully functional AI database and eventually a one-stop-shop AI data platform. EDB is entirely committed to make Postgres the natural choice for modern AI applications.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5519/Gianni Ciolli & Torsten Steinbach13:4500:45Ballsaal 1Sichere dein pgBackRest gegen Ransomware!pgBackRest ist aus guten Gründen sehr beliebt, unter anderem aufgrund seiner problemlosen Einbindung von Object Storage wie S3.
Es glänzt außerdem mit hoher Sicherheit, wenn man bei der Einrichtung den Schritten in der wirklich exzellenten Dokumentation folgt.
Trotzdem gibt es einige Dinge, die einem vielleicht nicht unbedingt sofort in's Auge springen, aber mit geringem Aufwand die Sicherheit noch einmal erhöhen.
Dieser Talk stellt einfache, aber sehr wirksame Maßnahmen vor, die den Schutz vor Datendiebstahl oder -verlust (z.B. durch Verschlüsselungstrojaner) signifikant verbessern, wenn S3 o.ä. zum Einsatz kommt.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5353/Gunnar "Nick" Bluth13:4500:45Ballsaal 2Professional PostgreSQL monitoring made easyThe talk firstly introduces all pertinent levels of database monitoring and then focuses on PostgreSQL and the means it provides. The meaning and importance of key metrics will be explained. As the Postgres community has already developed a lot of tools in that area, some popular common options will be highlighted together with the problems that different monitoring approaches have. To overcome some of these problems an Open Source tool from Cybertec, called pgwatch, is introduced and explained in detail to offer the simplest possibly entry into exhaustive Postgres monitoring. Also discussed will be advanced topics like anomaly detection and alerting, which can be easily implemented on top of the underlying data tier (TimescaleDB) with the help of the TICK stack or Grafana.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5266/Pavlo Golub13:4500:45Ballsaal 3Understanding PostgreSQL statistics to optimize performanceIn this session, I will explain the types of statistics in PostgreSQL - 1. Data Distribution statistics (which are generated when we run 'Analyze' and will also talk about 'Create Statistics' command) 2. Monitoring Statistics and its uses (most commonly used stat views - pg_stat_user_tables, pg_stat_user_indexes, pg_stat_bgwriter etc.)
The session will aim to explain these statistics and how to make use of them to optimize your workloads.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5369/Divya Sharma13:4500:45Sponsor TrackWhat enterprises need & want in Azure Database for PostgreSQLChoosing a managed service for your PostgreSQL database is an important decision. This session dives into managed service features in Azure Database for PostgreSQL that enterprises (and developers) have told us they need: High Availability (HA), Disaster Recovery (DR), Performance, Security, Authentication, Scalability, Backup & Recovery, and the list goes on. And of course you’ll learn about advanced monitoring and observability tools—plus how you can leverage AI capabilities in Azure Database for PostgreSQL via pgvector and azure_ai. In addition, you’ll learn how we integrate PostgreSQL on Azure with the latest releases of Postgres. Come learn why Azure Database for PostgreSQL – Flexible Server is becoming more and more popular in the enterprise.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5518/Silvano Coriani14:4000:45Ballsaal 1PostgreSQL Replikation: 15 Tücken und LösungenReplikation ist überall: wir benutzen es, um Redundanz, Hochverfügbarkeit, Aggregierung oder Verteilung von Daten, Datenbankmigrationen und Upgrades, oder sogar horizontalen Scale-Out umzusetzen. Zwar ist Replikation in PostgreSQL gut unterstützt und leicht umzusetzen, aber es gibt einige Missverständnisse, Ausnahmefälle und andere Probleme die über den Erfolg eines Projektes entscheiden können.
Dieser Vortrag befasst sich in erster Linie mit dem binären Replikationsmechanismus der in PostgreSQL eingebaut ist, wobei einige der Probleme auch für logische Repliaktion relevant sind. Wir werden häufig auftretende Probleme beleuchten, wie WAL-Recycling oder lang laufende Transaktionen, die in jedem Replikations-Setup bedacht werden müssen. Wir werden auch komplexere Probleme behandeln, z.B. Subtransaction-Overflow, die die Replikation verhindern oder sogar in Ausfallzeit resultieren können, wenn gewisse Umstände zusammenkommen. Insgesamt werden wir mindestens 15 verschiedene Probleme und Lösungen besprechen.
Ich installiere, verwalte und suche Fehler in PostgreSQL Replikations-Setups seit über fünf Jahren für große und kleine Kunden, die verschiedene Lasten und Anwendungen auf ihren Datenbanken betreiben.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5492/Julian Markwort14:4000:45Ballsaal 2Optimizing Your Cloud-Based PostgreSQL Setup: A Guide to PaaS Best Practices and Common PitfallsNavigating the complexities of a managed PostgreSQL instance on the cloud can often appear overwhelming, especially with the plethora of options and configurations available. As more organizations shift towards Platform as a Service (PaaS) for their database needs, understanding key best practices becomes imperative. This talk is aimed at demystifying the dos and don'ts of running a PostgreSQL instance in a PaaS model, with particular attention to Azure Database for PostgreSQL Flexible Server.
What You Will Learn:
1. What is Managed PostgreSQL (PaaS): An overview of what it means to operate a PostgreSQL database in a PaaS environment, highlighting the pros and cons.
2. Latency Concerns: The importance of application and database co-location in the same region and availability zone to minimize latency.
3. Logging Limitations: The potential performance drawbacks of indiscriminate logging and guidance on what to log and what to skip.
4. Choosing the Right SKU: Understanding the implications of your SKU selection and how it can impact scalability and operations in the future.
5. Operating Without Superuser: Understand why superuser access is restricted in PaaS environments and learn how to effectively manage your database operations without it.
Although the case studies and examples will focus on Azure Database for PostgreSQL Flexible Server, the principles and recommendations are applicable across other cloud providers.
Whether you’re a database administrator, developer, or someone curious about managed PostgreSQL databases, this talk aims to provide you with practical advice for maintaining an effective and efficient PaaS database environment.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5367/Alicja Kucharczyk14:4000:45Ballsaal 3Navigating the PostgreSQL Seas: A Quest To Hidden TreasuresIn this thrilling presentation, we will hoist the sails and navigate the tumultuous seas of PostgreSQL, the renowned open-source relational database. As intrepid developers, you are invited to embark on a quest to discover often overlooked features, the true hidden treasures of the database world. Join us as we explore the depths of unwavering transactional processing, chart the islands of complex data (including JSON and geographical data), and dive into the abyss of highly efficient text search techniques. By revealing these buried gems, you can not only optimize your applications but also become true programming pirates, ready to conquer new technological horizons. So, raise the black flag of PostgreSQL and prepare to set sail on this adventure in search of lost treasures, Arrr!https://www.postgresql.eu/events/pgconfde2024/schedule/session/5444/Lætitia AVROT15:2500:30Ballsaal 1Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5513/15:2500:30Ballsaal 2Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5514/15:2500:30Ballsaal 3Breakhttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5515/15:5500:50Ballsaal 1PostgreSQL 17 highlight: In-core inkrementelle Backups, Backup Ketten und einen Haufen DetailsVon Vielen schon lange gewünscht: Ab Version 17 wird PostgreSQL aller Wahrscheinlichkeit nach in der Lage sein, inkrementelle Backups zu erstellen, und diese natürlich auch wieder hierzustellen. In diesem Vortrag werden wir uns ansehen, wie inkrementelle Backups funktionieren, was es mit Backup Ketten auf sich hat und wie sich das in einem Backup-Konzept zusammenfügen liesse. Das werden wir uns alles auf der Konsole mit Bespielen erarbeiten, ohne Folien. Natürlich wird es aber einen Foliensatz geben, der das alles Schritt für Schritt erklärt, damit es dann von jedem nachvollzogen werden kann.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5489/Daniel Westermann15:5500:50Ballsaal 2Box Framework - Out of the Box web applications on top of PostgreSQL DatabaseBox is an open-source framework that aims to allow developers and DBA to quickly create and customize complex web applications on top of PostgreSQL databases, minimizing development and management costs. It is conceived to leave most of the logic and the user access rights on the database side, where it also stores the application configuration and the structure of the forms and pages in a dedicated schema. It can therefore take advantage of PostgreSQL advanced features like Row Level Security, updatable views, triggers, functions, etc. Box reads the DDL and builds up a complete web application with standard forms for each table. It easily allows the creation of custom forms and views, combining prebuilt widgets for each field type and more specialized ones like maps to handle GIS data, file managers with embedded image handling, nested tables, and more. This project is the result of a fruitful collaboration between the Swiss federal institute of forest, snow and landscape research WSL (www.wsl.ch) and wavein.ch. It has originally been developed to allow environmental researchers with database skills to manage their own research data, rapidly build web applications on top of it, e.g. to directly perform relevés in the field, export data or plot graphs. It is now becoming more and more feature rich and is moving outside the academic world, finding application in many other fields. During the talk, I will explain the framework's key features, the strengths, and why you may be interested in it. Finally, we will demo how fast it is to bootstrap a simple web application on a PostgreSQL database.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5476/Andrea Minetti15:5500:50Ballsaal 3Things your explain plan is not telling youAs much as we all love PostgreSQL explain command for diagnosing performance issues, there are some things that it doesn't reveal to us. In this talk I will take a look at some of the hidden performance gremlins and how to spot them in your plans. Things like TOAST lookups, visibility checks, hint bits and expression evaluation time can sometimes turn up in surprising places in your execution plans with little to no indication that this is happening. Knowing how to look for these problems is already a large step towards solving them.https://www.postgresql.eu/events/pgconfde2024/schedule/session/5466/Ants Aasma16:5000:20Ballsaal 1Closinghttps://www.postgresql.eu/events/pgconfde2024/schedule/session/5516/Andreas Scherbaum