PGConf.Russia 2025

This talk will likely be of interest to network providers who can enhance their cool features by adding something to the syntax to simplify administration, as well as anyone interested in machine learning. We will explore the SQL parsing mechanism, what, where, and how to modify it in order to introduce new syntax constructs. Additionally, I will demonstrate the capabilities of machine learning on tabular data, all within the database, using the new syntax.

t’s well known that pg_upgrade is the go-to tool for fast PostgreSQL upgrades. However, even with this tool, there are cases where the upgrade process takes far longer than expected.

In our case, upgrading a PostgreSQL database with 350,000 tables meant either waiting 3.5 hours or finding a better approach. By digging into the pg_upgrade source code, we discovered a way to speed up the process significantly. In this talk, we’ll share how we did it.

In PostgreSQL performance testing, benchmarks measure query execution time (latency). To get more reliable results, queries are executed repeatedly, generating a dataset of latency values. Performance is often assessed using standard metrics like the median or mean, but we propose a more advanced approach.

In practice, latency distributions are often multimodal, consisting of multiple underlying distributions with distinct characteristics. In such cases, traditional statistical methods are insufficient, requiring a more detailed analysis of the dataset’s structure.

Our work presents a tool that automatically performs statistical analysis of benchmark results, accounting for dataset-specific features. It detects multimodality, identifies the number and boundaries of dominant modes, and determines key distribution parameters—providing deeper insights into PostgreSQL performance variations.

What does a modern application developer need to transition from familiar cloud-based solutions to SQL databases?

This talk is a practical report on how we transformed an SQL database into a NoSQL-like solution with a developer-friendly interface. We’ll discuss:

What application developers expect from a data framework

How to implement declarative DDL, access control, and automation for reading and writing data

Simplified DML operations, pagination, and other essential features for application development