OPTIONS and HEAD requests are supported, both of which can use the functions supported by the Flask framework by default.
- Query inference server information.
1GET /v2
or1GET /v2/
Parameter
Description
Data Type
serverName
Inference service name
String
Response example:1{'errorCode': 0, 'errorMsg': 'Succeeded!', 'isSuccess': true, 'outputs': [{'server_name': 'StreamServer'}]}
- Check whether the inference server is running.
1GET /v2/live
or1GET /v2/live/
Parameter
Description
Data Type
isLive
Indicates whether the inference service is running.
Boolean
Response example:1{'errorCode': 0, 'errorMsg': 'Succeeded!', 'isSuccess': true, 'outputs': [{'isLive': True}]}
- Check whether the inference server is ready to accept inference requests.
GET /v2/ready
orGET /v2/ready/
Parameter
Description
Data Type
isReady
Indicates whether the inference server is ready to accept inference requests.
Boolean
- Check whether a model or stream is ready to accept inference requests.
GET v2/streams/${STREAM_NAME}/ready
GET v2/models/${MODEL_NAME}/readyParameter
Description
Data Type
isReady
Indicates whether a model or stream is ready to accept inference requests.
Boolean
Response example:1{'errorCode': 0, 'errorMsg': 'Succeeded!', 'isSuccess': true, 'outputs': [{'isReady': True}]}
${MODEL_NAME} and ${STREAM_NAME} must be character strings, which consist of letters, digits, and special characters (+-_).
- Query the configuration of a model or stream.
GET v2/streams/${STREAM_NAME}/config
GET v2/models/${MODEL_NAME}/configParameter
Description
Data Type
streamsConfig/modelsConfig
Inference configuration of a model or stream
JSON string
Response example:1{'errorCode': 0, 'errorMsg': 'Succeeded!', 'isSuccess': true, 'outputs': [{'ModelSample': {'deviceId': 1, 'dynamicBatching': {'dynamicStrategy': 'Nearest', 'preferredBatchSize': [1, 2, 4, 8], 'singleBatchInfer': 0, 'waitingTime': 100}, 'inferType': 'models', 'inputs': [{'dataType': 'UINT8', 'format': 'FORMAT_NHWC', 'id': 0, 'name': 'Placeholder', 'shape': [-1, 224, 224, 3]}], 'name': 'resnet50', 'outputs': [{'dataType': 'FLOAT32', 'format': 'FORMAT_NONE', 'id': 0, 'name': 'fp32_vars/dense/BiasAdd:0', 'shape': [-1, 1001]}], 'timeoutMs': 3000}}]}
${MODEL_NAME} and ${STREAM_NAME} must be character strings, which consist of letters, digits, and special characters (+-_).