Health Probe API

Checks whether an inference process is normal.

You are advised to call this API every 5 seconds.
This API supports only the prefill-decode hybrid deployment and does not support the prefill-decode disaggregation scenario.

Operation type: GET

URL: https://{ip}:{port}/health/timed-${TIMEOUT}

The {ip} field prioritizes reading the environment variable MIES_CONTAINER_MANAGEMENT_IP. If the environment variable does not exist, managementIpAddress in the configuration file is used. If managementIpAddress does not exist in the configuration file, ipAddress in the configuration file is used.
The {port} field prioritizes reading the managementPort parameter in the configuration file. If the configuration file does not contain the managementPort parameter, the port parameter in the configuration file is used.
{TIMEOUT} specifies the probe timeout interval. The value is an integer in the range of [1, 600], in seconds. If the value is not set, the default value 10 is used.
The MindIE health probe sends a request to the service. If the request is returned within ${TIMEOUT} seconds, the service is healthy. When the service is under heavy load, the MindIE health probe may time out. In such cases, the MindIE health probe should not be used as the sole indicator of whether the service is running properly. For details about how to configure status determination parameters, see Parameters in HealthCheckConfig.

None

Request example 1:

GET https://{ip}:{port}/health/timed

Request example 2:

GET https://{ip}:{port}/health/timed-200

Response example:

{
    "status": "healthy"
}

Response status code:

Status code 200: The service status is normal, and the message body contains no content.
Other status codes: The service status is abnormal.

Return Value	Type	Description
status	String	Health status.

Parent topic: Internal Endpoint APIs