Advanced PostgreSQL and PgBouncer metrics exporter for Prometheus with declarative configuration and 600+ metrics.
PG Exporter is an advanced metrics exporter that collects performance and health data from PostgreSQL databases and PgBouncer connection pools for Prometheus. It solves the problem of fragmented PostgreSQL monitoring by providing a unified, highly customizable solution with over 600 metrics out of the box, covering everything from basic uptime to detailed query analysis and replication lag.
Database administrators, DevOps engineers, and SREs who need comprehensive, production-grade monitoring for PostgreSQL clusters and PgBouncer instances, especially those managing self-hosted or on-premise deployments.
Developers choose PG Exporter for its declarative YAML configuration that allows deep customization of metrics, its battle-tested reliability in large-scale environments, and its unique ability to monitor both PostgreSQL and PgBouncer in a single tool with intelligent auto-discovery and dynamic query planning.
Advanced PostgreSQL & Pgbouncer Metrics Exporter for Prometheus
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Allows fine-grained control over metrics through customizable YAML files, enabling teams to define exactly what data to collect based on their specific needs, as shown in the extensive config examples.
Provides over 600 metrics covering PostgreSQL and PgBouncer, including extensions like TimescaleDB and Citus, offering deep observability out of the box for production environments.
Supports multiple query branches based on server conditions such as version, role, and extensions, ensuring compatibility and efficient scraping without manual adjustments.
Battle-tested in real-world environments for over six years across thousands of cores, ensuring stability and performance in large-scale deployments as highlighted in the README.
The extensive YAML config files, with over 600 lines of examples, can be overwhelming for new users and require significant time to master, potentially slowing down initial setup.
Assumes local-first deployment with SSL disabled by default, which could lead to security vulnerabilities if not explicitly configured for remote or sensitive environments, as noted in the connection URL defaults.
Can generate thousands of time series per instance, potentially slowing down scrape times for databases with many tables, necessitating manual tuning to avoid timeouts, as warned in the collector notes.