API Health Monitor

An API health monitoring service in Rust. Users register endpoints to monitor; a background scheduler checks them on their own intervals, records latency and status history, and sends webhook alerts when an endpoint's health status changes.

Stack: Rust · Axum · SQLx · PostgreSQL · Tokio · reqwest · JWT

How it works

A background task ticks every 5 seconds and queries the database for endpoints that are due for a check. This is computed in SQL by comparing each endpoint's last check time plus its check_interval_seconds against now(). Due endpoints are checked concurrently (5 at a time), each result is written to health_check, and if an endpoint's status changed since the last check, the user's active webhooks are notified.

Health statuses: Healthy (2xx response), Unhealthy (non-2xx response), Down (unreachable).

Setup

Requires Rust, PostgreSQL, and sqlx-cli.

cp .env.example .env     # set DATABASE_URL and JWT_SECRET
sqlx database create
sqlx migrate run         # required — migrations don't run automatically
cargo run

Server listens on 0.0.0.0:3000.

API

All routes except /register and /login require Authorization: Bearer <token>. Tokens expire after 1 hour.

Method	Path	Description
POST	/register	Create account — { email, password } (min 8 chars)
POST	/login	Get JWT — { email, password }
POST	/endpoints	Add endpoint to monitor
GET	/endpoints	List your endpoints
GET	/endpoints/{id}	Get one endpoint
PUT	/endpoints/{id}	Update endpoint (any subset of fields)
DELETE	/endpoints/{id}	Delete endpoint
GET	/endpoints/{id}/history	Check history for an endpoint
POST	/webhooks	Add webhook — { target_url, is_active? }
GET	/webhooks	List your webhooks
GET	/webhooks/{id}	Get one webhook
PUT	/webhooks/{id}	Update webhook
DELETE	/webhooks/{id}	Delete webhook

Creating an endpoint:

{
  "name": "My API",
  "url": "https://api.example.com/health",
  "check_interval_seconds": 60,
  "expected_status_code": 200,
  "is_active": true
}

name and url are required; the rest default to 60, 200, and true.

Webhook payload

Sent to every active webhook you own whenever one of your endpoints changes status (including its first check):

{
  "url": "https://api.example.com/health",
  "name": "My API",
  "status": 500,
  "health_status": "Unhealthy",
  "error_message": "Returned error code: 500 Internal Server Error"
}

status is 0 if the endpoint was unreachable; error_message is null when healthy.