mlrun 1.3.2rc1__py3-none-any.whl → 1.3.2rc2__py3-none-any.whl

mlrun/utils/model_monitoring.py

@@ -13,16 +13,97 @@
 # limitations under the License.
 #
 
-import json
-import warnings
-from typing import Union
+import enum
+import hashlib
+from dataclasses import dataclass
+from typing import Optional, Union
 
 import mlrun
 import mlrun.model
 import mlrun.model_monitoring.constants as model_monitoring_constants
 import mlrun.platforms.iguazio
+import mlrun.utils
 from mlrun.api.schemas.schedule import ScheduleCronTrigger
-from mlrun.config import is_running_as_api
+
+
+@dataclass
+class FunctionURI:
+    project: str
+    function: str
+    tag: Optional[str] = None
+    hash_key: Optional[str] = None
+
+    @classmethod
+    def from_string(cls, function_uri):
+        project, uri, tag, hash_key = mlrun.utils.parse_versioned_object_uri(
+            function_uri
+        )
+        return cls(
+            project=project,
+            function=uri,
+            tag=tag or None,
+            hash_key=hash_key or None,
+        )
+
+
+@dataclass
+class VersionedModel:
+    model: str
+    version: Optional[str]
+
+    @classmethod
+    def from_string(cls, model):
+        try:
+            model, version = model.split(":")
+        except ValueError:
+            model, version = model, None
+
+        return cls(model, version)
+
+
+@dataclass
+class EndpointUID:
+    project: str
+    function: str
+    function_tag: str
+    function_hash_key: str
+    model: str
+    model_version: str
+    uid: Optional[str] = None
+
+    def __post_init__(self):
+        function_ref = (
+            f"{self.function}_{self.function_tag or self.function_hash_key or 'N/A'}"
+        )
+        versioned_model = f"{self.model}_{self.model_version or 'N/A'}"
+        unique_string = f"{self.project}_{function_ref}_{versioned_model}"
+        self.uid = hashlib.sha1(unique_string.encode("utf-8")).hexdigest()
+
+    def __str__(self):
+        return self.uid
+
+
+def create_model_endpoint_id(function_uri: str, versioned_model: str):
+    function_uri = FunctionURI.from_string(function_uri)
+    versioned_model = VersionedModel.from_string(versioned_model)
+
+    if (
+        not function_uri.project
+        or not function_uri.function
+        or not versioned_model.model
+    ):
+        raise ValueError("Both function_uri and versioned_model have to be initialized")
+
+    uid = EndpointUID(
+        function_uri.project,
+        function_uri.function,
+        function_uri.tag,
+        function_uri.hash_key,
+        versioned_model.model,
+        versioned_model.version,
+    )
+
+    return uid
 
 
 def parse_model_endpoint_project_prefix(path: str, project_name: str):
@@ -35,20 +116,29 @@ def parse_model_endpoint_store_prefix(store_prefix: str):
     return endpoint, container, path
 
 
-def set_project_model_monitoring_credentials(access_key: str, project: str = None):
+def set_project_model_monitoring_credentials(
+    access_key: str, project: Optional[str] = None
+):
     """Set the credentials that will be used by the project's model monitoring
     infrastructure functions.
     The supplied credentials must have data access
+
     :param access_key: Model Monitoring access key for managing user permissions.
     :param project: The name of the model monitoring project.
     """
     mlrun.get_run_db().create_project_secrets(
         project=project or mlrun.mlconf.default_project,
         provider=mlrun.api.schemas.SecretProviderName.kubernetes,
-        secrets={model_monitoring_constants.ProjectSecretKeys.ACCESS_KEY: access_key},
+        secrets={"MODEL_MONITORING_ACCESS_KEY": access_key},
     )
 
 
+class EndpointType(enum.IntEnum):
+    NODE_EP = 1  # end point that is not a child of a router
+    ROUTER = 2  # endpoint that is router
+    LEAF_EP = 3  # end point that is a child of a router
+
+
 class TrackingPolicy(mlrun.model.ModelObj):
     """
     Modified model monitoring configurations. By using TrackingPolicy, the user can apply his model monitoring
@@ -125,119 +215,3 @@ class TrackingPolicy(mlrun.model.ModelObj):
                 model_monitoring_constants.EventFieldType.DEFAULT_BATCH_INTERVALS
             ] = self.default_batch_intervals.dict()
         return struct
-
-
-def get_connection_string(project: str = None):
-    """Get endpoint store connection string from the project secret.
-    If wasn't set, take it from the system configurations"""
-    if is_running_as_api():
-        # Running on API server side
-        import mlrun.api.crud.secrets
-        import mlrun.api.schemas
-
-        return (
-            mlrun.api.crud.secrets.Secrets().get_project_secret(
-                project=project,
-                provider=mlrun.api.schemas.secret.SecretProviderName.kubernetes,
-                allow_secrets_from_k8s=True,
-                secret_key=model_monitoring_constants.ProjectSecretKeys.ENDPOINT_STORE_CONNECTION,
-            )
-            or mlrun.mlconf.model_endpoint_monitoring.endpoint_store_connection
-        )
-    else:
-        # Running on stream server side
-        import mlrun
-
-        return (
-            mlrun.get_secret_or_env(
-                model_monitoring_constants.ProjectSecretKeys.ENDPOINT_STORE_CONNECTION
-            )
-            or mlrun.mlconf.model_endpoint_monitoring.endpoint_store_connection
-        )
-
-
-def get_stream_path(project: str = None):
-    # TODO: This function (as well as other methods in this file) includes both client and server side code. We will
-    #  need to refactor and adjust this file in the future.
-    """Get stream path from the project secret. If wasn't set, take it from the system configurations"""
-
-    if is_running_as_api():
-
-        # Running on API server side
-        import mlrun.api.crud.secrets
-        import mlrun.api.schemas
-
-        stream_uri = mlrun.api.crud.secrets.Secrets().get_project_secret(
-            project=project,
-            provider=mlrun.api.schemas.secret.SecretProviderName.kubernetes,
-            allow_secrets_from_k8s=True,
-            secret_key=model_monitoring_constants.ProjectSecretKeys.STREAM_PATH,
-        ) or mlrun.mlconf.get_model_monitoring_file_target_path(
-            project=project,
-            kind=model_monitoring_constants.FileTargetKind.STREAM,
-            target="online",
-        )
-
-    else:
-
-        import mlrun
-
-        stream_uri = mlrun.get_secret_or_env(
-            model_monitoring_constants.ProjectSecretKeys.STREAM_PATH
-        ) or mlrun.mlconf.get_model_monitoring_file_target_path(
-            project=project,
-            kind=model_monitoring_constants.FileTargetKind.STREAM,
-            target="online",
-        )
-
-    if stream_uri.startswith("kafka://"):
-        if "?topic" in stream_uri:
-            raise mlrun.errors.MLRunInvalidArgumentError(
-                "Custom kafka topic is not allowed"
-            )
-        # Add topic to stream kafka uri
-        stream_uri += f"?topic=monitoring_stream_{project}"
-
-    elif stream_uri.startswith("v3io://") and mlrun.mlconf.is_ce_mode():
-        # V3IO is not supported in CE mode, generating a default http stream path
-        stream_uri = mlrun.mlconf.model_endpoint_monitoring.default_http_sink
-
-    return stream_uri
-
-
-def validate_errors_and_metrics(endpoint: dict):
-    """
-    Replace default null values for `error_count` and `metrics` for users that logged a model endpoint before 1.3.0
-
-    Leaving here for backwards compatibility which related to the model endpoint schema
-
-    :param endpoint: An endpoint flattened dictionary.
-    """
-    warnings.warn(
-        "This will be deprecated in 1.3.0, and will be removed in 1.5.0",
-        # TODO: In 1.3.0 do changes in examples & demos In 1.5.0 remove
-        FutureWarning,
-    )
-
-    # Validate default value for `error_count`
-    # For backwards compatibility reasons, we validate that the model endpoint includes the `error_count` key
-    if (
-        model_monitoring_constants.EventFieldType.ERROR_COUNT in endpoint
-        and endpoint[model_monitoring_constants.EventFieldType.ERROR_COUNT] == "null"
-    ):
-        endpoint[model_monitoring_constants.EventFieldType.ERROR_COUNT] = "0"
-
-    # Validate default value for `metrics`
-    # For backwards compatibility reasons, we validate that the model endpoint includes the `metrics` key
-    if (
-        model_monitoring_constants.EventFieldType.METRICS in endpoint
-        and endpoint[model_monitoring_constants.EventFieldType.METRICS] == "null"
-    ):
-        endpoint[model_monitoring_constants.EventFieldType.METRICS] = json.dumps(
-            {
-                model_monitoring_constants.EventKeyMetrics.GENERIC: {
-                    model_monitoring_constants.EventLiveStats.LATENCY_AVG_1H: 0,
-                    model_monitoring_constants.EventLiveStats.PREDICTIONS_PER_SECOND: 0,
-                }
-            }
-        )

mlrun/utils/version/version.json

@@ -1,4 +1,4 @@
 {
-  "git_commit": "b95ec3c5e7e8954875e4a1eb82ef96e8222b7ce9",
-  "version": "1.3.2-rc1"
+  "git_commit": "c39a103bc6462422660bcdf40270b35469ad1b48",
+  "version": "1.3.2-rc2"
 }

{mlrun-1.3.2rc1.dist-info → mlrun-1.3.2rc2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mlrun
-Version: 1.3.2rc1
+Version: 1.3.2rc2
 Summary: Tracking and config of machine learning runs
 Home-page: https://github.com/mlrun/mlrun
 Author: Yaron Haviv
@@ -51,9 +51,9 @@ Requires-Dist: distributed (~=2021.11.2)
 Requires-Dist: kubernetes (~=12.0)
 Requires-Dist: humanfriendly (~=8.2)
 Requires-Dist: fastapi (~=0.92.0)
-Requires-Dist: fsspec (~=2023.1.0)
+Requires-Dist: fsspec (~=2021.8.1)
 Requires-Dist: v3iofs (~=0.1.15)
-Requires-Dist: storey (~=1.3.19)
+Requires-Dist: storey (~=1.3.15)
 Requires-Dist: deepdiff (~=5.0)
 Requires-Dist: pymysql (~=1.0)
 Requires-Dist: inflection (~=0.5.0)
@@ -61,15 +61,16 @@ Requires-Dist: python-dotenv (~=0.17.0)
 Requires-Dist: setuptools (~=65.5)
 Requires-Dist: deprecated (~=1.2)
 Provides-Extra: all
-Requires-Dist: adlfs (~=2022.2.0) ; extra == 'all'
-Requires-Dist: aiobotocore (~=2.4.2) ; extra == 'all'
+Requires-Dist: adlfs (~=2021.8.1) ; extra == 'all'
+Requires-Dist: aiobotocore (~=1.4.0) ; extra == 'all'
 Requires-Dist: azure-core (~=1.24) ; extra == 'all'
 Requires-Dist: azure-identity (~=1.5) ; extra == 'all'
 Requires-Dist: azure-keyvault-secrets (~=4.2) ; extra == 'all'
 Requires-Dist: azure-storage-blob (~=12.13) ; extra == 'all'
 Requires-Dist: bokeh (>=2.4.2,~=2.4) ; extra == 'all'
-Requires-Dist: boto3 (~=1.24.59) ; extra == 'all'
-Requires-Dist: gcsfs (~=2023.1.0) ; extra == 'all'
+Requires-Dist: boto3 (<1.17.107,~=1.9) ; extra == 'all'
+Requires-Dist: botocore (<1.20.107,>=1.20.106) ; extra == 'all'
+Requires-Dist: gcsfs (~=2021.8.1) ; extra == 'all'
 Requires-Dist: google-cloud-bigquery[bqstorage,pandas] (~=3.2) ; extra == 'all'
 Requires-Dist: google-cloud-storage (~=1.20) ; extra == 'all'
 Requires-Dist: google-cloud (~=0.34) ; extra == 'all'
@@ -79,7 +80,7 @@ Requires-Dist: msrest (~=0.6.21) ; extra == 'all'
 Requires-Dist: plotly (<5.12.0,~=5.4) ; extra == 'all'
 Requires-Dist: pyopenssl (>=23) ; extra == 'all'
 Requires-Dist: redis (~=4.3) ; extra == 'all'
-Requires-Dist: s3fs (~=2023.1.0) ; extra == 'all'
+Requires-Dist: s3fs (~=2021.8.1) ; extra == 'all'
 Provides-Extra: api
 Requires-Dist: uvicorn (~=0.20.0) ; extra == 'api'
 Requires-Dist: dask-kubernetes (~=0.11.0) ; extra == 'api'
@@ -90,7 +91,7 @@ Provides-Extra: azure-blob-storage
 Requires-Dist: msrest (~=0.6.21) ; extra == 'azure-blob-storage'
 Requires-Dist: azure-core (~=1.24) ; extra == 'azure-blob-storage'
 Requires-Dist: azure-storage-blob (~=12.13) ; extra == 'azure-blob-storage'
-Requires-Dist: adlfs (~=2022.2.0) ; extra == 'azure-blob-storage'
+Requires-Dist: adlfs (~=2021.8.1) ; extra == 'azure-blob-storage'
 Requires-Dist: pyopenssl (>=23) ; extra == 'azure-blob-storage'
 Provides-Extra: azure-key-vault
 Requires-Dist: azure-identity (~=1.5) ; extra == 'azure-key-vault'
@@ -99,14 +100,15 @@ Requires-Dist: pyopenssl (>=23) ; extra == 'azure-key-vault'
 Provides-Extra: bokeh
 Requires-Dist: bokeh (>=2.4.2,~=2.4) ; extra == 'bokeh'
 Provides-Extra: complete
-Requires-Dist: adlfs (~=2022.2.0) ; extra == 'complete'
-Requires-Dist: aiobotocore (~=2.4.2) ; extra == 'complete'
+Requires-Dist: adlfs (~=2021.8.1) ; extra == 'complete'
+Requires-Dist: aiobotocore (~=1.4.0) ; extra == 'complete'
 Requires-Dist: azure-core (~=1.24) ; extra == 'complete'
 Requires-Dist: azure-identity (~=1.5) ; extra == 'complete'
 Requires-Dist: azure-keyvault-secrets (~=4.2) ; extra == 'complete'
 Requires-Dist: azure-storage-blob (~=12.13) ; extra == 'complete'
-Requires-Dist: boto3 (~=1.24.59) ; extra == 'complete'
-Requires-Dist: gcsfs (~=2023.1.0) ; extra == 'complete'
+Requires-Dist: boto3 (<1.17.107,~=1.9) ; extra == 'complete'
+Requires-Dist: botocore (<1.20.107,>=1.20.106) ; extra == 'complete'
+Requires-Dist: gcsfs (~=2021.8.1) ; extra == 'complete'
 Requires-Dist: google-cloud-bigquery[bqstorage,pandas] (~=3.2) ; extra == 'complete'
 Requires-Dist: graphviz (~=0.20.0) ; extra == 'complete'
 Requires-Dist: kafka-python (~=2.0) ; extra == 'complete'
@@ -114,18 +116,19 @@ Requires-Dist: msrest (~=0.6.21) ; extra == 'complete'
 Requires-Dist: plotly (<5.12.0,~=5.4) ; extra == 'complete'
 Requires-Dist: pyopenssl (>=23) ; extra == 'complete'
 Requires-Dist: redis (~=4.3) ; extra == 'complete'
-Requires-Dist: s3fs (~=2023.1.0) ; extra == 'complete'
+Requires-Dist: s3fs (~=2021.8.1) ; extra == 'complete'
 Provides-Extra: complete-api
-Requires-Dist: adlfs (~=2022.2.0) ; extra == 'complete-api'
-Requires-Dist: aiobotocore (~=2.4.2) ; extra == 'complete-api'
+Requires-Dist: adlfs (~=2021.8.1) ; extra == 'complete-api'
+Requires-Dist: aiobotocore (~=1.4.0) ; extra == 'complete-api'
 Requires-Dist: apscheduler (~=3.6) ; extra == 'complete-api'
 Requires-Dist: azure-core (~=1.24) ; extra == 'complete-api'
 Requires-Dist: azure-identity (~=1.5) ; extra == 'complete-api'
 Requires-Dist: azure-keyvault-secrets (~=4.2) ; extra == 'complete-api'
 Requires-Dist: azure-storage-blob (~=12.13) ; extra == 'complete-api'
-Requires-Dist: boto3 (~=1.24.59) ; extra == 'complete-api'
+Requires-Dist: boto3 (<1.17.107,~=1.9) ; extra == 'complete-api'
+Requires-Dist: botocore (<1.20.107,>=1.20.106) ; extra == 'complete-api'
 Requires-Dist: dask-kubernetes (~=0.11.0) ; extra == 'complete-api'
-Requires-Dist: gcsfs (~=2023.1.0) ; extra == 'complete-api'
+Requires-Dist: gcsfs (~=2021.8.1) ; extra == 'complete-api'
 Requires-Dist: google-cloud-bigquery[bqstorage,pandas] (~=3.2) ; extra == 'complete-api'
 Requires-Dist: graphviz (~=0.20.0) ; extra == 'complete-api'
 Requires-Dist: kafka-python (~=2.0) ; extra == 'complete-api'
@@ -134,7 +137,7 @@ Requires-Dist: objgraph (~=3.5) ; extra == 'complete-api'
 Requires-Dist: plotly (<5.12.0,~=5.4) ; extra == 'complete-api'
 Requires-Dist: pyopenssl (>=23) ; extra == 'complete-api'
 Requires-Dist: redis (~=4.3) ; extra == 'complete-api'
-Requires-Dist: s3fs (~=2023.1.0) ; extra == 'complete-api'
+Requires-Dist: s3fs (~=2021.8.1) ; extra == 'complete-api'
 Requires-Dist: sqlite3-to-mysql (~=1.4) ; extra == 'complete-api'
 Requires-Dist: uvicorn (~=0.20.0) ; extra == 'complete-api'
 Provides-Extra: google-cloud
@@ -144,7 +147,7 @@ Requires-Dist: google-cloud (~=0.34) ; extra == 'google-cloud'
 Provides-Extra: google-cloud-bigquery
 Requires-Dist: google-cloud-bigquery[bqstorage,pandas] (~=3.2) ; extra == 'google-cloud-bigquery'
 Provides-Extra: google-cloud-storage
-Requires-Dist: gcsfs (~=2023.1.0) ; extra == 'google-cloud-storage'
+Requires-Dist: gcsfs (~=2021.8.1) ; extra == 'google-cloud-storage'
 Provides-Extra: graphviz
 Requires-Dist: graphviz (~=0.20.0) ; extra == 'graphviz'
 Provides-Extra: kafka
@@ -154,9 +157,10 @@ Requires-Dist: plotly (<5.12.0,~=5.4) ; extra == 'plotly'
 Provides-Extra: redis
 Requires-Dist: redis (~=4.3) ; extra == 'redis'
 Provides-Extra: s3
-Requires-Dist: boto3 (~=1.24.59) ; extra == 's3'
-Requires-Dist: aiobotocore (~=2.4.2) ; extra == 's3'
-Requires-Dist: s3fs (~=2023.1.0) ; extra == 's3'
+Requires-Dist: boto3 (<1.17.107,~=1.9) ; extra == 's3'
+Requires-Dist: botocore (<1.20.107,>=1.20.106) ; extra == 's3'
+Requires-Dist: aiobotocore (~=1.4.0) ; extra == 's3'
+Requires-Dist: s3fs (~=2021.8.1) ; extra == 's3'
 
 <a id="top"></a>
 [![Build Status](https://github.com/mlrun/mlrun/workflows/CI/badge.svg)](https://github.com/mlrun/mlrun/actions)