Healthcheck Endpoint

[AEstrellaS]

Adding a healthcheck endpoint to keep track of UWAPI's and its dependent backend services statuses.

• /health :

  • Returns the current health and status of the UWAPI middleware and its dependent backend services.

    UWAPI health fields:

    • status: (String) Always "ok". Indicates that UWAPI is online.
    • http_code: (Integer) Always HTTP 200. Indicates that UWAPI is online.
    • timestamp: (Integer) ISO 8601 UTC timestamp of when the health response was made (e.g., "2025-05-16T20:25:55Z").
    • uptime: (String) Time UWAPI has been running in Xd Yh Zm Ws format.
    • cpu_usage: (String) The percentage of CPU currently used (e.g., "21.2%").
    • memory_usage: (String) The percentage of total system memory used (e.g., "10.1%").
    • process_count: (String) Number of processes currently running on the system.
    • backend_services: (Object) An object describing the status of each backend service UWAPI depends on.

    For each available backend service the following health fields are appended backend_services:

    • status: (String) "ok" if the service responded successfully; "fail" if it did not respond or returned an error.
    • http_code: (Integer) Service health check response HTTP code.
    • latency: (Integer) Present only if the health check succeeded. Round-trip latency in milliseconds to reach the service.

    Below is an example response from /health:

    {
      "backend_services": {
        "gamesmanclassic": {
          "http_code": 200,
          "latency_ms": 109,
          "status": "ok"
        },
        "gamesmanone": {
          "error": "HTTPConnectionPool(host='nyc.cs.berkeley.edu', port=8084): Max retries exceeded with url: /health (Caused by ConnectTimeoutError(\u003Curllib3.connection.HTTPConnection object at 0x106b46c10\u003E, 'Connection to nyc.cs.berkeley.edu timed out. (connect timeout=1.5)'))",
          "status": "fail"
        },
        "gamesmanpuzzles": {
          "http_code": 200,
          "latency_ms": 40,
          "status": "ok"
        }
      },
      "cpu_usage": "6.1%",
      "memory_usage": "58.5%",
      "process_count": 887,
      "status": "ok",
      "timestamp": "2025-05-16T20:47:47Z",
      "uptime": "0d 0h 0m 1s"
    }

[AEstrellaS]

Healthcheck endpoint now in branch ae/health-endpoint.

[AEstrellaS]

The error field has been removed just in case it exposes sensitive information. The lack of the error field wouldn't really affect debugging since the healthcheck response includes a timestamp and the servers have a logger.