Conference Schedule - PGCon 2023

Modern PostgreSQL monitoring

Date: 2023-06-02
Time: 09:00–12:00
Room: DMS 1150
Level: Intermediate

In this tutorial, we'll discuss:

  • what every good Postgres monitoring must have
  • what infrastructure engineers and application developers usually expect from a database monitoring
  • how database engineers perform incident troubleshooting and what should be present in "Dashboard #1" (and corresponding "Runbook #1" for SREs and DBAs/DBREs) – a shallow but wide look at parts of PostgreSQL and its various aspects, to quickly identify in which areas we need to dive deeper
  • why query analysis should be the central part of any database monitoring and what we can have in 2023 if we use only open source components: pg_stat_statements, pg_stat_kcache, pg_wait_sampling, and log analysis
  • very young yet very promising path: use of eBPF to monitor various database aspects, including query analysis

We'll also discuss the current state of monitoring systems – both open source and commercial – and how to get the most out of them.

As a result, I want to share my 18+ years of experience working with Postgres and have everyone who attends this tutorial to: - know how to close the gap quickly, if the current monitoring is unsatisfactory or there is no Postgres monitoring at all, and - khow how to build a great solution in the longer term, if there is such a goal.

There is no video for this session.


Nikolay Samokhvalov