pgwatch

pgwatch collects and displays metrics from PostgreSQL databases. A remote daemon periodically connects to each database and reads its statistics, using a normal database user without special rights or server-side extensions. Grafana presents the metrics and supports additional graphs; optional alerts can be sent through Mail, Slack, or PagerDuty. Metrics can be extended with pure SQL, and the listed storage choices are PostgreSQL, Prometheus, JSON File, and RPC. A web interface manages monitored databases and metric definitions and displays logs. pgwatch supports PostgreSQL 14 and newer, including single instances, high-availability clusters, pooled environments using PgBouncer or Pgpool II, and Patroni-managed clusters. It is free, open-source software for self-hosted deployment. The product page describes a one-minute Docker installation, with custom installations also possible. It is intended for large infrastructures and can be deployed fully on premises. Some operating-system metrics may be unavailable on cloud-managed PostgreSQL services because of provider restrictions.

pgwatch combines PostgreSQL metrics collection with Grafana visualization and a web management interface. Its free, self-hosted model and support for varied PostgreSQL deployments are useful, though cloud platform restrictions can limit some operating-system metrics.

pgwatch suits teams monitoring PostgreSQL, from individual instances to large infrastructures and high-availability or pooled deployments. It may also suit organizations that need on-premises deployment for data-residency requirements.

pgwatch is a free, open-source monitoring system for PostgreSQL teams that prefer to run their own tooling. It is most suitable for operators managing varied PostgreSQL topologies who need extensible metrics and database-level analysis without installing extensions on each server. The trade-off is that managed cloud platforms can withhold some operating-system data.

Overview

pgwatch uses a remote daemon that periodically connects to PostgreSQL and reads its statistics. It operates as a regular database user, requiring neither elevated privileges nor extensions on the monitored database server. That low-impact model is useful when production teams want monitoring without changing database installations.

PostgreSQL 14 and newer are supported. The deployment scope includes standalone databases, high-availability clusters, PgBouncer and Pgpool II pooled environments, and Patroni-managed clusters. CYBERTEC positions the software for large infrastructures, including environments with thousands of systems. It can also be deployed entirely on premises, which suits organizations with strict data-residency requirements.

Key features

Collection and database analysis

Support for query analysis, wait and lock analysis, resource metrics, and query-plan capture gives operators coverage from workload behavior to contention and capacity signals. Because collection is periodic, pgwatch is designed for continuing operational visibility rather than replacing PostgreSQL's own diagnostic tools.

Dashboards, alerts, and storage

Grafana provides the presentation layer, and teams can create additional graphs instead of being confined to fixed views. Optional Grafana alerting can deliver notifications through Mail, Slack, or PagerDuty. Metrics may be stored in PostgreSQL, Prometheus, JSON files, or RPC, allowing the monitoring design to follow an existing stack.

Extensibility and administration

Custom metrics are defined in pure SQL, either through the configuration database or YAML files. This is a meaningful advantage for teams with PostgreSQL expertise: application-specific measurements can be added without waiting for a vendor integration. The built-in web interface manages monitored databases and metrics, controls preset definitions, and shows logs, concentrating routine administration in one place.

Installation is advertised as a one-minute Docker deployment, while custom installations are also supported. The approach is flexible, but self-hosting still leaves your organization responsible for operating the monitoring service and its Grafana and storage components.

Pricing

There are no paid seats, usage quotas, or renewal charges stated for the Open Source plan. Its cost advantage is strongest for teams able to run and maintain self-hosted infrastructure. CYBERTEC also offers support packages: bug fixes for extensions are included in Standard, Professional, and Enterprise tiers, while ticket-system access is listed for Professional and Enterprise. Those packages are relevant when production operations require vendor-backed assistance rather than community or internal support.

Platforms

pgwatch is self-hosted and accessed through the web. It monitors PostgreSQL 14 and newer across single instances, high-availability clusters, PgBouncer or Pgpool II pools, and Patroni clusters. Cloud-managed PostgreSQL remains a qualification point: provider restrictions can make some operating-system-level metrics unavailable. Organizations needing complete infrastructure telemetry should confirm what their cloud service exposes before committing.

Who it's for

Choose pgwatch if your team operates PostgreSQL directly, needs to monitor many systems or deployment patterns, and values SQL-defined customization. Its normal-user collection model is particularly suitable where database-server changes and special privileges are tightly controlled. On-premises deployment also fits residency-sensitive environments.

Look elsewhere if you want a fully managed monitoring service or depend on operating-system metrics that a cloud provider does not expose. Self-hosting, Grafana, storage, and alert delivery remain part of your operational responsibility.

Pros and cons

Pros

• Low-privilege collection: a regular database account is sufficient, with no server extensions required.
• Broad topology coverage: single instances, HA, poolers, and Patroni are supported.
• Deep customization: pure-SQL metrics and user-created Grafana graphs adapt monitoring to local workloads.
• Open-source price: the Open Source plan costs nothing.
• Operational scale: the maker targets infrastructures containing thousands of systems.

Cons

• Cloud visibility gaps: managed services may block some operating-system metrics.
• Self-hosting work: your team must run the daemon, web interface, Grafana, and chosen storage.
• Database scope: support is for PostgreSQL 14 and newer, not a general multi-database monitor.

Alternatives

Site24x7 Database Monitoring is a paid, multi-platform service with a free plan, free trial, and a Lite tier at $9 a month when billed annually; choose it when hosted access across web, mobile, Linux, and Windows matters more than self-hosted control.

AimBetter suits buyers seeking paid per-server monitoring, starting at $110 a month, rather than an open-source deployment.

MySQL is the better fit when the monitored estate is MySQL-based or needs its own replication and high-availability components.

Grafana Cloud Application Observability is preferable when you want Grafana delivered as a cloud service, with a free tier of 2,232 host hours and usage-based Pro pricing.

Amazon RDS fits teams seeking managed relational databases and AWS Free Tier credits rather than a self-hosted PostgreSQL monitoring layer.

DBmarlin is an alternative freemium, self-hosted and web monitoring product with a free Starter plan.

Percona Monitoring and Management is another free, open-source database monitoring and observability option, with installation infrastructure potentially costing extra.

AI SQL Tuner Studio is aimed at Windows users tuning SQL Server Developer Edition, not broad PostgreSQL infrastructure monitoring.

Database Performance Monitoring Software provides a broader category comparison when pgwatch's PostgreSQL focus is too narrow.

Verdict

pgwatch is a strong recommendation for technically capable teams running PostgreSQL on infrastructure they control. Its zero-price open-source model, low-privilege collector, topology coverage, and SQL extensibility make it a practical choice for large or customized estates. Look elsewhere when you need a managed service, cross-database coverage, or guaranteed operating-system metrics from a cloud-hosted PostgreSQL provider.