Using TimescaleDB and Collectd for observability

This article discusses a minimal observability stack using TimescaleDB, Collectd, and Grafana for system metrics. It provides a production setup example, highlighting the benefits of SQL-powered analytics, flexible retention, and automatic compression. To replicate the setup, install PostgreSQL 14+, TimescaleDB, and the TimescaleDB Toolkit extensions, as well as Collectd 5.x with the PostgreSQL plugin. Create a dedicated Collectd user and database, and apply the provided schema to normalize metric identifiers.