def health_status(
user_model: Any, seldon_metrics: SeldonMetrics
) -> Union[prediction_pb2.SeldonMessage, List, Dict]:
"""
Call the user model to check the health of the model
Parameters
----------
user_model
User defined class instance
Returns
-------
Health check output
"""
if hasattr(user_model, "health_status_raw"):
try:
return user_model.health_status_raw()
except SeldonNotImplementedError:
pass
client_response = client_health_status(user_model)
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response_json(user_model, False, {}, client_response,
None, metrics)
def test_metrics_clear():
metrics = SeldonMetrics()
metrics.update([RAW_COUNTER_METRIC], method="predict")
assert len(metrics.data) > 0
metrics.clear()
assert len(metrics.data) == 0
def client_custom_metrics(
user_model: SeldonComponent,
seldon_metrics: SeldonMetrics,
method: str,
runtime_metrics: List[Dict] = [],
) -> List[Dict]:
"""
Get custom metrics for client and update SeldonMetrics.
This function will return empty list if INCLUDE_METRICS_IN_CLIENT_RESPONSE environmental
variable is NOT set to "true" or "True".
Parameters
----------
user_model
A Seldon user model
seldon_metrics
A SeldonMetrics instance
method:
tag of a method that collected the metrics
runtime_metrics:
metrics that were defined on runtime
Returns
-------
A list of custom metrics
"""
if not validate_metrics(runtime_metrics):
raise SeldonMicroserviceException(
f"Bad metric created during request: {json.dumps(runtime_metrics)}",
status_code=500,
reason="MICROSERVICE_BAD_METRIC",
)
seldon_metrics.update(runtime_metrics, method)
if hasattr(user_model, "metrics"):
try:
metrics = user_model.metrics()
if not validate_metrics(metrics):
raise SeldonMicroserviceException(
f"Bad metric created during request: {json.dumps(metrics)}",
status_code=500,
reason="MICROSERVICE_BAD_METRIC",
)
seldon_metrics.update(metrics, method)
if INCLUDE_METRICS_IN_CLIENT_RESPONSE:
return metrics + runtime_metrics
else:
return []
except SeldonNotImplementedError:
pass
logger.debug("custom_metrics is not implemented")
if INCLUDE_METRICS_IN_CLIENT_RESPONSE:
return runtime_metrics
else:
return []
def handle_raw_custom_metrics(
msg: Union[prediction_pb2.SeldonMessage, Dict],
seldon_metrics: SeldonMetrics,
is_proto: bool,
):
"""
Update SeldonMetrics object with custom metrics from raw methods.
If INCLUDE_METRICS_IN_CLIENT_RESPONSE environmental variable is set to "true"
metrics will be dropped from msg.
"""
if is_proto:
metrics = seldon_message_to_json(msg.meta).get("metrics", [])
if metrics and not INCLUDE_METRICS_IN_CLIENT_RESPONSE:
del msg.meta.metrics[:]
else:
metrics = msg.get("meta", {}).get("metrics", [])
if metrics and not INCLUDE_METRICS_IN_CLIENT_RESPONSE:
del msg["meta"]["metrics"]
seldon_metrics.update(metrics)
def client_custom_metrics(user_model: SeldonComponent,
seldon_metrics: SeldonMetrics) -> List[Dict]:
"""
Get custom metrics for client and update SeldonMetrics.
This function will return empty list if INCLUDE_METRICS_IN_CLIENT_RESPONSE environmental
variable is NOT set to "true" or "True".
Parameters
----------
user_model
A Seldon user model
seldon_metrics
A SeldonMetrics instance
Returns
-------
A list of custom metrics
"""
if hasattr(user_model, "metrics"):
try:
metrics = user_model.metrics()
if not validate_metrics(metrics):
j_str = json.dumps(metrics)
raise SeldonMicroserviceException(
"Bad metric created during request: " + j_str,
reason="MICROSERVICE_BAD_METRIC",
)
seldon_metrics.update(metrics)
if INCLUDE_METRICS_IN_CLIENT_RESPONSE:
return metrics
else:
return []
except SeldonNotImplementedError:
pass
logger.debug("custom_metrics is not implemented")
return []
def aggregate(
user_model: Any,
request: Union[prediction_pb2.SeldonMessageList, List, Dict],
seldon_metrics: SeldonMetrics,
) -> Union[prediction_pb2.SeldonMessage, List, Dict]:
"""
Aggregate a list of payloads
Parameters
----------
user_model
A Seldon user model
request
SeldonMessage proto
Returns
-------
Aggregated SeldonMessage proto
"""
def merge_meta(meta_list):
tags = {}
for meta in meta_list:
if meta:
tags.update(meta.get("tags", {}))
return {"tags": tags}
def merge_metrics(meta_list, custom_metrics):
metrics = []
for meta in meta_list:
if meta:
metrics.extend(meta.get("metrics", []))
metrics.extend(custom_metrics)
return metrics
is_proto = isinstance(request, prediction_pb2.SeldonMessageList)
if hasattr(user_model, "aggregate_rest"):
logger.warning(
"aggregate_rest is deprecated. Please use aggregate_raw")
return user_model.aggregate_rest(request)
elif hasattr(user_model, "aggregate_grpc"):
logger.warning(
"aggregate_grpc is deprecated. Please use aggregate_raw")
return user_model.aggregate_grpc(request)
else:
if hasattr(user_model, "aggregate_raw"):
try:
response = user_model.aggregate_raw(request)
if is_proto:
metrics = seldon_message_to_json(response.meta).get(
"metrics", [])
else:
metrics = response.get("meta", {}).get("metrics", [])
seldon_metrics.update(metrics)
return response
except SeldonNotImplementedError:
pass
if is_proto:
features_list = []
names_list = []
meta_list = []
for msg in request.seldonMessages:
(features, meta, datadef,
data_type) = extract_request_parts(msg)
features_list.append(features)
names_list.append(datadef.names)
meta_list.append(meta)
client_response = client_aggregate(user_model, features_list,
names_list)
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response(
user_model,
False,
request.seldonMessages[0],
client_response,
merge_meta(meta_list),
merge_metrics(meta_list, metrics),
)
else:
features_list = []
names_list = []
if isinstance(request, list):
msgs = request
elif "seldonMessages" in request and isinstance(
request["seldonMessages"], list):
msgs = request["seldonMessages"]
else:
raise SeldonMicroserviceException(
f"Invalid request data type: {request}")
meta_list = []
for msg in msgs:
(features, meta, datadef,
data_type) = extract_request_parts_json(msg)
class_names = datadef[
"names"] if datadef and "names" in datadef else []
features_list.append(features)
names_list.append(class_names)
meta_list.append(meta)
client_response = client_aggregate(user_model, features_list,
names_list)
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response_json(
user_model,
False,
msgs[0],
client_response,
merge_meta(meta_list),
merge_metrics(meta_list, metrics),
)
def predict(
user_model: Any,
request: Union[prediction_pb2.SeldonMessage, List, Dict],
seldon_metrics: SeldonMetrics,
) -> Union[prediction_pb2.SeldonMessage, List, Dict]:
"""
Call the user model to get a prediction and package the response
Parameters
----------
user_model
User defined class instance
request
The incoming request
Returns
-------
The prediction
"""
is_proto = isinstance(request, prediction_pb2.SeldonMessage)
if hasattr(user_model, "predict_rest") and not is_proto:
logger.warning("predict_rest is deprecated. Please use predict_raw")
return user_model.predict_rest(request)
elif hasattr(user_model, "predict_grpc") and is_proto:
logger.warning("predict_grpc is deprecated. Please use predict_raw")
return user_model.predict_grpc(request)
else:
if hasattr(user_model, "predict_raw"):
try:
response = user_model.predict_raw(request)
if is_proto:
metrics = seldon_message_to_json(response.meta).get(
"metrics", [])
else:
metrics = response.get("meta", {}).get("metrics", [])
seldon_metrics.update(metrics)
return response
except SeldonNotImplementedError:
pass
if is_proto:
(features, meta, datadef,
data_type) = extract_request_parts(request)
client_response = client_predict(user_model,
features,
datadef.names,
meta=meta)
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response(user_model, False, request,
client_response, meta, metrics)
else:
(features, meta, datadef,
data_type) = extract_request_parts_json(request)
class_names = datadef[
"names"] if datadef and "names" in datadef else []
client_response = client_predict(user_model,
features,
class_names,
meta=meta)
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response_json(user_model, False, request,
client_response, meta, metrics)
def route(
user_model: Any,
request: Union[prediction_pb2.SeldonMessage, List, Dict],
seldon_metrics: SeldonMetrics,
) -> Union[prediction_pb2.SeldonMessage, List, Dict]:
"""
Parameters
----------
user_model
A Seldon user model
request
A SelodonMessage proto
Returns
-------
"""
is_proto = isinstance(request, prediction_pb2.SeldonMessage)
if hasattr(user_model, "route_rest"):
logger.warning("route_rest is deprecated. Please use route_raw")
return user_model.route_rest(request)
elif hasattr(user_model, "route_grpc"):
logger.warning("route_grpc is deprecated. Please use route_raw")
return user_model.route_grpc(request)
else:
if hasattr(user_model, "route_raw"):
try:
response = user_model.route_raw(request)
if is_proto:
metrics = seldon_message_to_json(response.meta).get(
"metrics", [])
else:
metrics = response.get("meta", {}).get("metrics", [])
seldon_metrics.update(metrics)
return response
except SeldonNotImplementedError:
pass
if is_proto:
(features, meta, datadef,
data_type) = extract_request_parts(request)
client_response = client_route(user_model,
features,
datadef.names,
meta=meta)
if not isinstance(client_response, int):
raise SeldonMicroserviceException(
"Routing response must be int but got " +
str(client_response))
client_response_arr = np.array([[client_response]])
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response(user_model, False, request,
client_response_arr, None, metrics)
else:
(features, meta, datadef,
data_type) = extract_request_parts_json(request)
class_names = datadef[
"names"] if datadef and "names" in datadef else []
client_response = client_route(user_model,
features,
class_names,
meta=meta)
if not isinstance(client_response, int):
raise SeldonMicroserviceException(
"Routing response must be int but got " +
str(client_response))
client_response_arr = np.array([[client_response]])
metrics = client_custom_metrics(user_model)
if seldon_metrics is not None:
seldon_metrics.update(metrics)
return construct_response_json(user_model, False, request,
client_response_arr, None, metrics)
