mlrun 1.7.0rc16__py3-none-any.whl → 1.7.0rc18__py3-none-any.whl

mlrun/track/tracker.py

@@ -31,8 +31,9 @@ class Tracker(ABC):
     * Offline: Manually importing models and artifacts into an MLRun project using the `import_x` methods.
     """
 
+    @staticmethod
     @abstractmethod
-    def is_enabled(self) -> bool:
+    def is_enabled() -> bool:
         """
         Checks if tracker is enabled.
 

mlrun/utils/async_http.py

@@ -24,7 +24,7 @@ from aiohttp_retry import ExponentialRetry, RequestParams, RetryClient, RetryOpt
 from aiohttp_retry.client import _RequestContext
 
 from mlrun.config import config
-from mlrun.errors import err_to_str
+from mlrun.errors import err_to_str, raise_for_status
 
 from .helpers import logger as mlrun_logger
 
@@ -46,12 +46,21 @@ class AsyncClientWithRetry(RetryClient):
         *args,
         **kwargs,
     ):
+        # do not retry on PUT / PATCH as they might have side effects (not truly idempotent)
+        blacklisted_methods = (
+            blacklisted_methods
+            if blacklisted_methods is not None
+            else [
+                "POST",
+                "PUT",
+                "PATCH",
+            ]
+        )
         super().__init__(
             *args,
             retry_options=ExponentialRetryOverride(
                 retry_on_exception=retry_on_exception,
-                # do not retry on PUT / PATCH as they might have side effects (not truly idempotent)
-                blacklisted_methods=blacklisted_methods or ["POST", "PUT", "PATCH"],
+                blacklisted_methods=blacklisted_methods,
                 attempts=max_retries,
                 statuses=retry_on_status_codes,
                 factor=retry_backoff_factor,
@@ -63,6 +72,12 @@ class AsyncClientWithRetry(RetryClient):
             **kwargs,
         )
 
+    def methods_blacklist_update_required(self, new_blacklist: str):
+        self._retry_options: ExponentialRetryOverride
+        return set(self._retry_options.blacklisted_methods).difference(
+            set(new_blacklist)
+        )
+
     def _make_requests(
         self,
         params_list: list[RequestParams],
@@ -173,7 +188,7 @@ class _CustomRequestContext(_RequestContext):
                 last_attempt = current_attempt == self._retry_options.attempts
                 if self._is_status_code_ok(response.status) or last_attempt:
                     if self._raise_for_status:
-                        response.raise_for_status()
+                        raise_for_status(response)
 
                     self._response = response
                     return response
@@ -275,6 +290,11 @@ class _CustomRequestContext(_RequestContext):
                 if isinstance(exc.os_error, exc_type):
                     return
         if exc.__cause__:
-            return self.verify_exception_type(exc.__cause__)
+            # If the cause exception is retriable, return, otherwise, raise the original exception
+            try:
+                self.verify_exception_type(exc.__cause__)
+            except Exception:
+                raise exc
+            return
         else:
             raise exc

mlrun/utils/helpers.py

@@ -973,6 +973,15 @@ def get_ui_url(project, uid=None):
     return url
 
 
+def get_model_endpoint_url(project, model_name, model_endpoint_id):
+    url = ""
+    if mlrun.mlconf.resolve_ui_url():
+        url = f"{mlrun.mlconf.resolve_ui_url()}/{mlrun.mlconf.ui.projects_prefix}/{project}/models"
+        if model_name:
+            url += f"/model-endpoints/{model_name}/{model_endpoint_id}/overview"
+    return url
+
+
 def get_workflow_url(project, id=None):
     url = ""
     if mlrun.mlconf.resolve_ui_url():
@@ -1183,7 +1192,7 @@ def calculate_dataframe_hash(dataframe: pandas.DataFrame):
     return hashlib.sha1(pandas.util.hash_pandas_object(dataframe).values).hexdigest()
 
 
-def template_artifact_path(artifact_path, project, run_uid="project"):
+def template_artifact_path(artifact_path, project, run_uid=None):
     """
     Replace {{run.uid}} with the run uid and {{project}} with the project name in the artifact path.
     If no run uid is provided, the word `project` will be used instead as it is assumed to be a project
@@ -1191,6 +1200,7 @@ def template_artifact_path(artifact_path, project, run_uid="project"):
     """
     if not artifact_path:
         return artifact_path
+    run_uid = run_uid or "project"
     artifact_path = artifact_path.replace("{{run.uid}}", run_uid)
     artifact_path = _fill_project_path_template(artifact_path, project)
     return artifact_path
@@ -1572,13 +1582,19 @@ def validate_component_version_compatibility(
     component_current_version = None
     try:
         if component_name == "iguazio":
-            parsed_current_version = mlrun.mlconf.get_parsed_igz_version()
             component_current_version = mlrun.mlconf.igz_version
+            parsed_current_version = mlrun.mlconf.get_parsed_igz_version()
+
+            # ignore pre-release and build metadata, as iguazio version always has them, and we only care about the
+            # major, minor, and patch versions
+            parsed_current_version = semver.VersionInfo.parse(
+                f"{parsed_current_version.major}.{parsed_current_version.minor}.{parsed_current_version.patch}"
+            )
         if component_name == "nuclio":
+            component_current_version = mlrun.mlconf.nuclio_version
             parsed_current_version = semver.VersionInfo.parse(
                 mlrun.mlconf.nuclio_version
             )
-            component_current_version = mlrun.mlconf.nuclio_version
         if not parsed_current_version:
             return True
     except ValueError:
@@ -1597,3 +1613,12 @@ def validate_component_version_compatibility(
         if parsed_current_version < parsed_min_version:
             return False
     return True
+
+
+def format_alert_summary(
+    alert: mlrun.common.schemas.AlertConfig, event_data: mlrun.common.schemas.Event
+) -> str:
+    result = alert.summary.replace("{{project}}", alert.project)
+    result = result.replace("{{name}}", alert.name)
+    result = result.replace("{{entity}}", event_data.entity.ids[0])
+    return result

mlrun/utils/logger.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 import logging
+import typing
 from enum import Enum
 from sys import stdout
 from traceback import format_exception
@@ -221,11 +222,15 @@ class FormatterKinds(Enum):
     JSON = "json"
 
 
-def create_formatter_instance(formatter_kind: FormatterKinds) -> logging.Formatter:
+def resolve_formatter_by_kind(
+    formatter_kind: FormatterKinds,
+) -> type[
+    typing.Union[HumanReadableFormatter, HumanReadableExtendedFormatter, JSONFormatter]
+]:
     return {
-        FormatterKinds.HUMAN: HumanReadableFormatter(),
-        FormatterKinds.HUMAN_EXTENDED: HumanReadableExtendedFormatter(),
-        FormatterKinds.JSON: JSONFormatter(),
+        FormatterKinds.HUMAN: HumanReadableFormatter,
+        FormatterKinds.HUMAN_EXTENDED: HumanReadableExtendedFormatter,
+        FormatterKinds.JSON: JSONFormatter,
     }[formatter_kind]
 
 
@@ -243,11 +248,11 @@ def create_logger(
     logger_instance = Logger(level, name=name, propagate=False)
 
     # resolve formatter
-    formatter_instance = create_formatter_instance(
+    formatter_instance = resolve_formatter_by_kind(
         FormatterKinds(formatter_kind.lower())
     )
 
     # set handler
-    logger_instance.set_handler("default", stream or stdout, formatter_instance)
+    logger_instance.set_handler("default", stream or stdout, formatter_instance())
 
     return logger_instance

mlrun/utils/notifications/notification/slack.py

@@ -32,6 +32,7 @@ class SlackNotification(NotificationBase):
         "completed": ":smiley:",
         "running": ":man-running:",
         "error": ":x:",
+        "skipped": ":zzz:",
     }
 
     async def push(
@@ -135,8 +136,16 @@ class SlackNotification(NotificationBase):
         line = [
             self._get_slack_row(f":bell: {alert.name} alert has occurred"),
             self._get_slack_row(f"*Project:*\n{alert.project}"),
-            self._get_slack_row(f"*UID:*\n{event_data.entity.ids[0]}"),
+            self._get_slack_row(f"*ID:*\n{event_data.entity.ids[0]}"),
         ]
+
+        if alert.summary:
+            line.append(
+                self._get_slack_row(
+                    f"*Summary:*\n{mlrun.utils.helpers.format_alert_summary(alert, event_data)}"
+                )
+            )
+
         if event_data.value_dict:
             data_lines = []
             for key, value in event_data.value_dict.items():
@@ -144,10 +153,21 @@ class SlackNotification(NotificationBase):
             data_text = "\n".join(data_lines)
             line.append(self._get_slack_row(f"*Event data:*\n{data_text}"))
 
-        if url := mlrun.utils.helpers.get_ui_url(
-            alert.project, event_data.entity.ids[0]
-        ):
-            line.append(self._get_slack_row(f"*Overview:*\n<{url}|*Job overview*>"))
+        if (
+            event_data.entity.kind == mlrun.common.schemas.alert.EventEntityKind.JOB
+        ):  # JOB entity
+            uid = event_data.value_dict.get("uid")
+            url = mlrun.utils.helpers.get_ui_url(alert.project, uid)
+            overview_type = "Job overview"
+        else:  # MODEL entity
+            model_name = event_data.value_dict.get("model")
+            model_endpoint_id = event_data.value_dict.get("model_endpoint_id")
+            url = mlrun.utils.helpers.get_model_endpoint_url(
+                alert.project, model_name, model_endpoint_id
+            )
+            overview_type = "Model endpoint"
+
+        line.append(self._get_slack_row(f"*Overview:*\n<{url}|*{overview_type}*>"))
 
         return line
 
@@ -157,11 +177,11 @@ class SlackNotification(NotificationBase):
 
         # Only show the URL if the run is not a function (serving or mlrun function)
         kind = run.get("step_kind")
-        if url and not kind or kind == "run":
+        state = run["status"].get("state", "")
+        if state != "skipped" and (url and not kind or kind == "run"):
             line = f'<{url}|*{meta.get("name")}*>'
         else:
             line = meta.get("name")
-        state = run["status"].get("state", "")
         if kind:
             line = f'{line} *({run.get("step_kind", run.get("kind", ""))})*'
         line = f'{self.emojis.get(state, ":question:")}  {line}'

mlrun/utils/notifications/notification_pusher.py

@@ -14,7 +14,6 @@
 
 import asyncio
 import datetime
-import json
 import os
 import re
 import traceback
@@ -23,6 +22,7 @@ from concurrent.futures import ThreadPoolExecutor
 
 import kfp
 import mlrun_pipelines.common.ops
+import mlrun_pipelines.models
 
 import mlrun.common.runtimes.constants
 import mlrun.common.schemas
@@ -392,17 +392,29 @@ class NotificationPusher(_NotificationPusherBase):
         steps = []
         db = mlrun.get_run_db()
 
-        def _add_run_step(_node_name, _node_template, _step_kind):
-            _run = db.list_runs(
-                project=run.metadata.project,
-                labels=f"mlrun/runner-pod={_node_name}",
-            )[0]
-            _run["step_kind"] = _step_kind
+        def _add_run_step(_step: mlrun_pipelines.models.PipelineStep):
+            try:
+                _run = db.list_runs(
+                    project=run.metadata.project,
+                    labels=f"mlrun/runner-pod={_step.node_name}",
+                )[0]
+            except IndexError:
+                _run = {
+                    "metadata": {
+                        "name": _step.display_name,
+                        "project": run.metadata.project,
+                    },
+                }
+            _run["step_kind"] = _step.step_type
+            if _step.skipped:
+                _run.setdefault("status", {})["state"] = (
+                    mlrun.common.runtimes.constants.RunStates.skipped
+                )
             steps.append(_run)
 
-        def _add_deploy_function_step(_, _node_template, _step_kind):
+        def _add_deploy_function_step(_step: mlrun_pipelines.models.PipelineStep):
             project, name, hash_key = self._extract_function_uri(
-                _node_template["metadata"]["annotations"]["mlrun/function-uri"]
+                _step.get_annotation("mlrun/function-uri")
             )
             if name:
                 try:
@@ -419,16 +431,19 @@ class NotificationPusher(_NotificationPusherBase):
                             "hash_key": hash_key,
                         },
                     }
-                function["status"] = {
-                    "state": mlrun.common.runtimes.constants.PodPhases.pod_phase_to_run_state(
-                        node["phase"]
-                    ),
-                }
+                pod_phase = _step.phase
+                if _step.skipped:
+                    state = mlrun.common.schemas.FunctionState.skipped
+                else:
+                    state = mlrun.common.runtimes.constants.PodPhases.pod_phase_to_run_state(
+                        pod_phase
+                    )
+                function["status"] = {"state": state}
                 if isinstance(function["metadata"].get("updated"), datetime.datetime):
                     function["metadata"]["updated"] = function["metadata"][
                         "updated"
                     ].isoformat()
-                function["step_kind"] = _step_kind
+                function["step_kind"] = _step.step_type
                 steps.append(function)
 
         step_methods = {
@@ -446,26 +461,10 @@ class NotificationPusher(_NotificationPusherBase):
             return steps
 
         try:
-            workflow_nodes = sorted(
-                workflow_manifest["status"]["nodes"].items(),
-                key=lambda _node: _node[1]["finishedAt"],
-            )
-            for node_name, node in workflow_nodes:
-                if node["type"] != "Pod":
-                    # Skip the parent DAG node
-                    continue
-
-                node_template = next(
-                    template
-                    for template in workflow_manifest["spec"]["templates"]
-                    if template["name"] == node["templateName"]
-                )
-                step_type = node_template["metadata"]["annotations"].get(
-                    "mlrun/pipeline-step-type"
-                )
-                step_method = step_methods.get(step_type)
+            for step in workflow_manifest.get_steps():
+                step_method = step_methods.get(step.step_type)
                 if step_method:
-                    step_method(node_name, node_template, step_type)
+                    step_method(step)
             return steps
         except Exception:
             # If we fail to read the pipeline steps, we will return the list of runs that have the same workflow id
@@ -481,19 +480,24 @@ class NotificationPusher(_NotificationPusherBase):
             )
 
     @staticmethod
-    def _get_workflow_manifest(workflow_id: str) -> typing.Optional[dict]:
-        kfp_client = kfp.Client(namespace=mlrun.mlconf.namespace)
+    def _get_workflow_manifest(
+        workflow_id: str,
+    ) -> typing.Optional[mlrun_pipelines.models.PipelineManifest]:
+        kfp_url = mlrun.mlconf.resolve_kfp_url(mlrun.mlconf.namespace)
+        if not kfp_url:
+            raise mlrun.errors.MLRunNotFoundError(
+                "KubeFlow Pipelines is not configured"
+            )
+
+        kfp_client = kfp.Client(host=kfp_url)
 
         # arbitrary timeout of 5 seconds, the workflow should be done by now
         kfp_run = kfp_client.wait_for_run_completion(workflow_id, 5)
         if not kfp_run:
             return None
 
-        kfp_run = kfp_run.to_dict()
-        try:
-            return json.loads(kfp_run["pipeline_runtime"]["workflow_manifest"])
-        except Exception:
-            return None
+        kfp_run = mlrun_pipelines.models.PipelineRun(kfp_run)
+        return kfp_run.workflow_manifest()
 
     def _extract_function_uri(self, function_uri: str) -> tuple[str, str, str]:
         """

mlrun/utils/version/version.json

@@ -1,4 +1,4 @@
 {
-  "git_commit": "db779a53a32a737e10d538e2fef25695a962117a",
-  "version": "1.7.0-rc16"
+  "git_commit": "cf983306a4f164f1c0a4f3ccf666ba9448d09e2e",
+  "version": "1.7.0-rc18"
 }

{mlrun-1.7.0rc16.dist-info → mlrun-1.7.0rc18.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: mlrun
-Version: 1.7.0rc16
+Version: 1.7.0rc18
 Summary: Tracking and config of machine learning runs
 Home-page: https://github.com/mlrun/mlrun
 Author: Yaron Haviv
@@ -28,7 +28,7 @@ Requires-Dist: aiohttp-retry ~=2.8
 Requires-Dist: click ~=8.1
 Requires-Dist: nest-asyncio ~=1.0
 Requires-Dist: ipython ~=8.10
-Requires-Dist: nuclio-jupyter ~=0.9.16
+Requires-Dist: nuclio-jupyter ~=0.9.17
 Requires-Dist: numpy <1.27.0,>=1.16.5
 Requires-Dist: pandas <2.2,>=1.2
 Requires-Dist: pyarrow <15,>=10.0
@@ -43,7 +43,7 @@ Requires-Dist: semver ~=3.0
 Requires-Dist: dependency-injector ~=4.41
 Requires-Dist: fsspec <2024.4,>=2023.9.2
 Requires-Dist: v3iofs ~=0.1.17
-Requires-Dist: storey ~=1.7.11
+Requires-Dist: storey ~=1.7.17
 Requires-Dist: inflection ~=0.5.0
 Requires-Dist: python-dotenv ~=0.17.0
 Requires-Dist: setuptools ~=69.1
@@ -82,6 +82,7 @@ Requires-Dist: pyopenssl >=23 ; extra == 'all'
 Requires-Dist: redis ~=4.3 ; extra == 'all'
 Requires-Dist: s3fs <2024.4,>=2023.9.2 ; extra == 'all'
 Requires-Dist: sqlalchemy ~=1.4 ; extra == 'all'
+Requires-Dist: taos-ws-py ~=0.3.2 ; extra == 'all'
 Provides-Extra: api
 Requires-Dist: uvicorn ~=0.27.1 ; extra == 'api'
 Requires-Dist: dask-kubernetes ~=0.11.0 ; extra == 'api'
@@ -129,6 +130,7 @@ Requires-Dist: pyopenssl >=23 ; extra == 'complete'
 Requires-Dist: redis ~=4.3 ; extra == 'complete'
 Requires-Dist: s3fs <2024.4,>=2023.9.2 ; extra == 'complete'
 Requires-Dist: sqlalchemy ~=1.4 ; extra == 'complete'
+Requires-Dist: taos-ws-py ~=0.3.2 ; extra == 'complete'
 Provides-Extra: complete-api
 Requires-Dist: adlfs ==2023.9.0 ; extra == 'complete-api'
 Requires-Dist: aiobotocore <2.8,>=2.5.0 ; extra == 'complete-api'
@@ -161,6 +163,7 @@ Requires-Dist: pyopenssl >=23 ; extra == 'complete-api'
 Requires-Dist: redis ~=4.3 ; extra == 'complete-api'
 Requires-Dist: s3fs <2024.4,>=2023.9.2 ; extra == 'complete-api'
 Requires-Dist: sqlalchemy ~=1.4 ; extra == 'complete-api'
+Requires-Dist: taos-ws-py ~=0.3.2 ; extra == 'complete-api'
 Requires-Dist: timelength ~=1.1 ; extra == 'complete-api'
 Requires-Dist: uvicorn ~=0.27.1 ; extra == 'complete-api'
 Provides-Extra: dask
@@ -193,6 +196,8 @@ Requires-Dist: aiobotocore <2.8,>=2.5.0 ; extra == 's3'
 Requires-Dist: s3fs <2024.4,>=2023.9.2 ; extra == 's3'
 Provides-Extra: sqlalchemy
 Requires-Dist: sqlalchemy ~=1.4 ; extra == 'sqlalchemy'
+Provides-Extra: tdengine
+Requires-Dist: taos-ws-py ~=0.3.2 ; extra == 'tdengine'
 
 <a id="top"></a>
 [![Build Status](https://github.com/mlrun/mlrun/actions/workflows/build.yaml/badge.svg?branch=development)](https://github.com/mlrun/mlrun/actions/workflows/build.yaml?query=branch%3Adevelopment)