Checkout the slides

Abstract:

In this session, we will touch on the importance of PostgreSQL monitoring, even when you are using managed offerings like RDS, CloudSQL, etc. I will then delve into how we can leverage open-source tools like postgres_exporter, Prometheus, and Grafana to effectively monitor PostgreSQL.

We will explore the key metrics and performance indicators to track in order to ensure optimal database performance and reliability. We will discuss practical examples and best practices for setting up and configuring the postgres_exporter to collect and expose relevant metrics.

Additionally, we will delve into the pg_hint_plan extension and auto_explain module, which helps in analyzing and optimizing query performance. The session will cover enabling and configuring these modules to capture query plans and execution details. Furthermore, we'll explore how to correlate these details with the exporter metrics for in-depth performance analysis.

Target Audience:

• PostgreSQL administrators
• Site Reliability Engineers

Key Takeaways:

• Understanding the importance of monitoring PostgreSQL databases.
• Knowledge of key metrics and performance indicators to track and alert on.
• Practical ways to hunt down slow queries.