Watchdog

Watchdog is the platform’s health and operability surface.

It gives operators and developers a single place to check whether the environment is alive and whether the major hosts are reachable.

Why Watchdog exists

In a modular platform, one green API process is not enough.

An environment also depends on:

• the API host;
• the Background Jobs host;
• the Watchdog host itself;
• the web client reachability path;
• external dependencies such as PostgreSQL and Keycloak where relevant.

Watchdog exists to make that health picture easy to inspect.

What Watchdog typically provides

A Watchdog host usually provides:

• a health endpoint;
• a health UI / aggregate view;
• configured checks for the relevant hosts;
• links back into the environment.

Local PM example

In the Property Management local Docker Compose environment, the Watchdog host aggregates checks for:

• itself;
• the Property Management API;
• the Property Management Background Jobs host.

The Watchdog UI is exposed at:

https://localhost:7075/health-ui

The base health endpoint is:

https://localhost:7075/health

Why a dedicated host is useful

A dedicated Watchdog surface avoids mixing operator diagnostics into the main product UI and gives a stable place to:

• verify environment readiness;
• troubleshoot startup order;
• confirm host-to-host reachability.

Startup dependency role

In the local Docker environment, Watchdog starts after the critical application pieces are available.

That ordering makes it a good final checkpoint that the whole environment is healthy enough for interactive work.

What good health checks look like

A good health check should be:

• cheap enough to run regularly;
• meaningful enough to detect a real problem;
• scoped to a dependency that matters;
• not so broad that it becomes noisy or flaky.

What Watchdog is not

Watchdog is not a substitute for:

• structured logs;
• traces;
• business integrity scans;
• alerting strategy.

It is a visibility surface, not the whole observability stack.

Developer guidance

When composing a new vertical Watchdog host:

• include the vertical’s API and background jobs;
• keep naming clear and environment-specific;
• expose a clean operator UI;
• avoid checks that are too expensive or too noisy;
• use it as a real operability surface, not a demo-only extra.

A platform with API, jobs, and migrations but no coherent health surface is much harder to operate confidently. Watchdog closes that gap.