prefect-client 3.3.5.dev2__tar.gz → 3.3.5.dev4__tar.gz

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prefect-client
-Version: 3.3.5.dev2
+Version: 3.3.5.dev4
 Summary: Workflow orchestration and management.
 Project-URL: Changelog, https://github.com/PrefectHQ/prefect/releases
 Project-URL: Documentation, https://docs.prefect.io

prefect_client-3.3.5.dev4/src/prefect/_build_info.py

@@ -0,0 +1,5 @@
+# Generated by versioningit
+__version__ = "3.3.5.dev4"
+__build_date__ = "2025-04-17 08:08:34.376039+00:00"
+__git_commit__ = "d90d684c9e0773b41bf092ad7ef17e22dda08c77"
+__dirty__ = False

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/_experimental/bundles.py

@@ -8,6 +8,7 @@ import multiprocessing.context
 import os
 import subprocess
 import sys
+from pathlib import Path
 from typing import Any, TypedDict
 
 import cloudpickle
@@ -99,7 +100,9 @@ def extract_flow_from_bundle(bundle: SerializedBundle) -> Flow[Any, Any]:
 
 
 def _extract_and_run_flow(
-    bundle: SerializedBundle, env: dict[str, Any] | None = None
+    bundle: SerializedBundle,
+    cwd: Path | str | None = None,
+    env: dict[str, Any] | None = None,
 ) -> None:
     """
     Extracts a flow from a bundle and runs it.
@@ -108,6 +111,7 @@ def _extract_and_run_flow(
 
     Args:
         bundle: The bundle to extract and run.
+        cwd: The working directory to use when running the flow.
         env: The environment to use when running the flow.
     """
 
@@ -120,6 +124,9 @@ def _extract_and_run_flow(
     context = _deserialize_bundle_object(bundle["context"])
     flow_run = FlowRun.model_validate(bundle["flow_run"])
 
+    if cwd:
+        os.chdir(cwd)
+
     with SettingsContext(
         profile=settings_context.profile,
         settings=Settings(),
@@ -138,6 +145,8 @@ def _extract_and_run_flow(
 
 def execute_bundle_in_subprocess(
     bundle: SerializedBundle,
+    env: dict[str, Any] | None = None,
+    cwd: Path | str | None = None,
 ) -> multiprocessing.context.SpawnProcess:
     """
     Executes a bundle in a subprocess.
@@ -150,6 +159,7 @@ def execute_bundle_in_subprocess(
     """
 
     ctx = multiprocessing.get_context("spawn")
+    env = env or {}
 
     # Install dependencies if necessary
     if dependencies := bundle.get("dependencies"):
@@ -164,7 +174,9 @@ def execute_bundle_in_subprocess(
         kwargs={
             "bundle": bundle,
             "env": get_current_settings().to_environment_variables(exclude_unset=True)
-            | os.environ,
+            | os.environ
+            | env,
+            "cwd": cwd,
         },
     )
 

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/events/clients.py

@@ -39,6 +39,7 @@ import prefect.types._datetime
 from prefect.events import Event
 from prefect.logging import get_logger
 from prefect.settings import (
+    PREFECT_API_AUTH_STRING,
     PREFECT_API_KEY,
     PREFECT_API_SSL_CERT_FILE,
     PREFECT_API_TLS_INSECURE_SKIP_VERIFY,
@@ -367,7 +368,7 @@ class PrefectEventsClient(EventsClient):
         await self._connect.__aexit__(exc_type, exc_val, exc_tb)
         return await super().__aexit__(exc_type, exc_val, exc_tb)
 
-    def _log_debug(self, message: str, *args, **kwargs) -> None:
+    def _log_debug(self, message: str, *args: Any, **kwargs: Any) -> None:
         message = f"EventsClient(id={id(self)}): " + message
         logger.debug(message, *args, **kwargs)
 
@@ -578,6 +579,7 @@ class PrefectEventSubscriber:
     _seen_events: MutableMapping[UUID, bool]
 
     _api_key: Optional[str]
+    _auth_token: Optional[str]
 
     def __init__(
         self,
@@ -593,6 +595,8 @@ class PrefectEventSubscriber:
                 the client should attempt to reconnect
         """
         self._api_key = None
+        self._auth_token = PREFECT_API_AUTH_STRING.value()
+
         if not api_url:
             api_url = cast(str, PREFECT_API_URL.value())
 
@@ -641,8 +645,10 @@ class PrefectEventSubscriber:
         await pong
 
         logger.debug("  authenticating...")
+        # Use the API key (for Cloud) OR the auth token (for self-hosted with auth string)
+        token = self._api_key or self._auth_token
         await self._websocket.send(
-            orjson.dumps({"type": "auth", "token": self._api_key}).decode()
+            orjson.dumps({"type": "auth", "token": token}).decode()
         )
 
         try:
@@ -652,13 +658,13 @@ class PrefectEventSubscriber:
         except AssertionError as e:
             raise Exception(
                 "Unable to authenticate to the event stream. Please ensure the "
-                "provided api_key you are using is valid for this environment. "
+                "provided api_key or auth_token you are using is valid for this environment. "
                 f"Reason: {e.args[0]}"
             )
         except ConnectionClosedError as e:
             reason = getattr(e.rcvd, "reason", None)
             msg = "Unable to authenticate to the event stream. Please ensure the "
-            msg += "provided api_key you are using is valid for this environment. "
+            msg += "provided api_key or auth_token you are using is valid for this environment. "
             msg += f"Reason: {reason}" if reason else ""
             raise Exception(msg) from e
 

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/flows.py

@@ -40,6 +40,7 @@ from typing import (
 from uuid import UUID
 
 import pydantic
+from exceptiongroup import BaseExceptionGroup, ExceptionGroup
 from pydantic.v1 import BaseModel as V1BaseModel
 from pydantic.v1.decorator import ValidatedFunction as V1ValidatedFunction
 from pydantic.v1.errors import ConfigError  # TODO
@@ -105,6 +106,9 @@ from ._internal.pydantic.v2_validated_func import (
     V2ValidatedFunction as ValidatedFunction,
 )
 
+if TYPE_CHECKING:
+    from prefect.workers.base import BaseWorker
+
 T = TypeVar("T")  # Generic type var for capturing the inner return type of async funcs
 R = TypeVar("R")  # The return type of the user's function
 P = ParamSpec("P")  # The parameters of the flow
@@ -1130,7 +1134,7 @@ class Flow(Generic[P, R]):
     @classmethod
     async def afrom_source(
         cls,
-        source: Union[str, "RunnerStorage", ReadableDeploymentStorage],
+        source: Union[str, Path, "RunnerStorage", ReadableDeploymentStorage],
         entrypoint: str,
     ) -> "Flow[..., Any]":
         """
@@ -1244,7 +1248,7 @@ class Flow(Generic[P, R]):
     @async_dispatch(afrom_source)
     def from_source(
         cls,
-        source: Union[str, "RunnerStorage", ReadableDeploymentStorage],
+        source: Union[str, Path, "RunnerStorage", ReadableDeploymentStorage],
         entrypoint: str,
     ) -> "Flow[..., Any]":
         """
@@ -1976,7 +1980,7 @@ class FlowDecorator:
         # manually here.
         @staticmethod
         def from_source(
-            source: Union[str, "RunnerStorage", ReadableDeploymentStorage],
+            source: Union[str, Path, "RunnerStorage", ReadableDeploymentStorage],
             entrypoint: str,
         ) -> Union["Flow[..., Any]", Coroutine[Any, Any, "Flow[..., Any]"]]: ...
 
@@ -1984,6 +1988,141 @@ class FlowDecorator:
 flow: FlowDecorator = FlowDecorator()
 
 
+class InfrastructureBoundFlow(Flow[P, R]):
+    """
+    EXPERIMENTAL: This class is experimental and may be removed or changed in future
+        releases.
+
+    A flow that is bound to running on a specific infrastructure.
+
+    Attributes:
+        work_pool: The name of the work pool to run the flow on. The base job
+            configuration of the work pool will determine the configuration of the
+            infrastructure the flow will run on.
+        job_variables: Infrastructure configuration that will override the base job
+            configuration of the work pool.
+        worker_cls: The class of the worker to use to spin up infrastructure and submit
+            the flow to it.
+    """
+
+    def __init__(
+        self,
+        *args: Any,
+        work_pool: str,
+        job_variables: dict[str, Any],
+        worker_cls: type["BaseWorker[Any, Any, Any]"],
+        **kwargs: Any,
+    ):
+        super().__init__(*args, **kwargs)
+        self.work_pool = work_pool
+        self.job_variables = job_variables
+        self.worker_cls = worker_cls
+
+    @overload
+    def __call__(self: "Flow[P, NoReturn]", *args: P.args, **kwargs: P.kwargs) -> None:
+        # `NoReturn` matches if a type can't be inferred for the function which stops a
+        # sync function from matching the `Coroutine` overload
+        ...
+
+    @overload
+    def __call__(
+        self: "Flow[P, Coroutine[Any, Any, T]]",
+        *args: P.args,
+        **kwargs: P.kwargs,
+    ) -> Coroutine[Any, Any, T]: ...
+
+    @overload
+    def __call__(
+        self: "Flow[P, T]",
+        *args: P.args,
+        **kwargs: P.kwargs,
+    ) -> T: ...
+
+    @overload
+    def __call__(
+        self: "Flow[P, Coroutine[Any, Any, T]]",
+        *args: P.args,
+        return_state: Literal[True],
+        **kwargs: P.kwargs,
+    ) -> Awaitable[State[T]]: ...
+
+    @overload
+    def __call__(
+        self: "Flow[P, T]",
+        *args: P.args,
+        return_state: Literal[True],
+        **kwargs: P.kwargs,
+    ) -> State[T]: ...
+
+    def __call__(
+        self,
+        *args: "P.args",
+        return_state: bool = False,
+        wait_for: Optional[Iterable[PrefectFuture[Any]]] = None,
+        **kwargs: "P.kwargs",
+    ):
+        async def modified_call(
+            *args: P.args,
+            return_state: bool = False,
+            # TODO: Handle wait_for once we have an asynchronous way to wait for futures
+            # We should wait locally for futures to resolve before spinning up
+            # infrastructure.
+            wait_for: Optional[Iterable[PrefectFuture[Any]]] = None,
+            **kwargs: P.kwargs,
+        ) -> R | State[R]:
+            try:
+                async with self.worker_cls(work_pool_name=self.work_pool) as worker:
+                    parameters = get_call_parameters(self, args, kwargs)
+                    future = await worker.submit(
+                        flow=self,
+                        parameters=parameters,
+                        job_variables=self.job_variables,
+                    )
+                    if return_state:
+                        await future.wait_async()
+                        return future.state
+                    return await future.aresult()
+            except (ExceptionGroup, BaseExceptionGroup) as exc:
+                # For less verbose tracebacks
+                exceptions = exc.exceptions
+                if len(exceptions) == 1:
+                    raise exceptions[0] from None
+                else:
+                    raise
+
+        if inspect.iscoroutinefunction(self.fn):
+            return modified_call(
+                *args, return_state=return_state, wait_for=wait_for, **kwargs
+            )
+        else:
+            return run_coro_as_sync(
+                modified_call(
+                    *args,
+                    return_state=return_state,
+                    wait_for=wait_for,
+                    **kwargs,
+                )
+            )
+
+
+def bind_flow_to_infrastructure(
+    flow: Flow[P, R],
+    work_pool: str,
+    worker_cls: type["BaseWorker[Any, Any, Any]"],
+    job_variables: dict[str, Any] | None = None,
+) -> InfrastructureBoundFlow[P, R]:
+    new = InfrastructureBoundFlow[P, R](
+        flow.fn,
+        work_pool=work_pool,
+        job_variables=job_variables or {},
+        worker_cls=worker_cls,
+    )
+    # Copy all attributes from the original flow
+    for attr, value in flow.__dict__.items():
+        setattr(new, attr, value)
+    return new
+
+
 def _raise_on_name_with_banned_characters(name: Optional[str]) -> Optional[str]:
     """
     Raise an InvalidNameError if the given name contains any invalid

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/runner/runner.py

@@ -636,7 +636,12 @@ class Runner:
 
                     return process
 
-    async def execute_bundle(self, bundle: SerializedBundle) -> None:
+    async def execute_bundle(
+        self,
+        bundle: SerializedBundle,
+        cwd: Path | str | None = None,
+        env: dict[str, str | None] | None = None,
+    ) -> None:
         """
         Executes a bundle in a subprocess.
         """
@@ -651,7 +656,7 @@ class Runner:
             if not self._acquire_limit_slot(flow_run.id):
                 return
 
-            process = execute_bundle_in_subprocess(bundle)
+            process = execute_bundle_in_subprocess(bundle, cwd=cwd, env=env)
 
             if process.pid is None:
                 # This shouldn't happen because `execute_bundle_in_subprocess` starts the process

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/telemetry/instrumentation.py

@@ -1,6 +1,6 @@
 import logging
-import os
 import re
+import socket
 from typing import TYPE_CHECKING
 from urllib.parse import urljoin
 from uuid import UUID
@@ -74,7 +74,7 @@ def setup_exporters(
     resource = Resource.create(
         {
             "service.name": "prefect",
-            "service.instance.id": os.uname().nodename,
+            "service.instance.id": socket.gethostname(),
             "prefect.account": str(account_id),
             "prefect.workspace": str(workspace_id),
         }

{prefect_client-3.3.5.dev2 → prefect_client-3.3.5.dev4}/src/prefect/workers/base.py

@@ -3,7 +3,11 @@ from __future__ import annotations
 import abc
 import asyncio
 import datetime
+import json
+import subprocess
+import tempfile
 import threading
+import uuid
 import warnings
 from contextlib import AsyncExitStack
 from functools import partial
@@ -21,6 +25,7 @@ from zoneinfo import ZoneInfo
 import anyio
 import anyio.abc
 import httpx
+from exceptiongroup import BaseExceptionGroup, ExceptionGroup
 from importlib_metadata import (
     distributions,  # type: ignore[reportUnknownVariableType] incomplete typing
 )
@@ -35,6 +40,7 @@ from prefect._internal.schemas.validators import return_v_or_none
 from prefect.client.base import ServerType
 from prefect.client.orchestration import PrefectClient, get_client
 from prefect.client.schemas.actions import WorkPoolCreate, WorkPoolUpdate
+from prefect.client.schemas.objects import Flow as APIFlow
 from prefect.client.schemas.objects import (
     Integration,
     StateType,
@@ -48,6 +54,7 @@ from prefect.exceptions import (
     Abort,
     ObjectNotFound,
 )
+from prefect.futures import PrefectFlowRunFuture
 from prefect.logging.loggers import (
     PrefectLogAdapter,
     flow_run_logger,
@@ -81,11 +88,12 @@ from prefect.utilities.templating import (
 from prefect.utilities.urls import url_for
 
 if TYPE_CHECKING:
-    from prefect.client.schemas.objects import Flow, FlowRun
+    from prefect.client.schemas.objects import FlowRun
     from prefect.client.schemas.responses import (
         DeploymentResponse,
         WorkerFlowRunResponse,
     )
+    from prefect.flows import Flow
 
 
 class BaseJobConfiguration(BaseModel):
@@ -217,7 +225,7 @@ class BaseJobConfiguration(BaseModel):
         self,
         flow_run: "FlowRun",
         deployment: "DeploymentResponse | None" = None,
-        flow: "Flow | None" = None,
+        flow: "APIFlow | None" = None,
         work_pool: "WorkPool | None" = None,
         worker_name: str | None = None,
     ) -> None:
@@ -314,7 +322,7 @@ class BaseJobConfiguration(BaseModel):
         return labels
 
     @staticmethod
-    def _base_flow_labels(flow: "Flow | None") -> dict[str, str]:
+    def _base_flow_labels(flow: "APIFlow | None") -> dict[str, str]:
         if flow is None:
             return {}
 
@@ -418,6 +426,7 @@ class BaseWorkerResult(BaseModel, abc.ABC):
 C = TypeVar("C", bound=BaseJobConfiguration)
 V = TypeVar("V", bound=BaseVariables)
 R = TypeVar("R", bound=BaseWorkerResult)
+FR = TypeVar("FR")  # used to capture the return type of a flow
 
 
 @register_base_type
@@ -685,6 +694,146 @@ class BaseWorker(abc.ABC, Generic[C, V, R]):
             "Workers must implement a method for running submitted flow runs"
         )
 
+    async def submit(
+        self,
+        flow: "Flow[..., FR]",
+        parameters: dict[str, Any] | None = None,
+        job_variables: dict[str, Any] | None = None,
+    ) -> "PrefectFlowRunFuture[FR]":
+        """
+        EXPERIMENTAL: The interface for this method is subject to change.
+
+        Submits a flow to run via the worker.
+
+        Args:
+            flow: The flow to submit
+            parameters: The parameters to pass to the flow
+
+        Returns:
+            A flow run object
+        """
+        warnings.warn(
+            "Ad-hoc flow submission via workers is experimental. The interface "
+            "and behavior of this feature are subject to change.",
+            category=FutureWarning,
+        )
+        if self._runs_task_group is None:
+            raise RuntimeError("Worker not properly initialized")
+
+        flow_run = await self._runs_task_group.start(
+            partial(
+                self._submit_adhoc_run,
+                flow=flow,
+                parameters=parameters,
+                job_variables=job_variables,
+            ),
+        )
+        return PrefectFlowRunFuture(flow_run_id=flow_run.id)
+
+    async def _submit_adhoc_run(
+        self,
+        flow: "Flow[..., FR]",
+        parameters: dict[str, Any] | None = None,
+        job_variables: dict[str, Any] | None = None,
+        task_status: anyio.abc.TaskStatus["FlowRun"] | None = None,
+    ):
+        """
+        Submits a flow run to the Kubernetes worker.
+        """
+        from prefect._experimental.bundles import (
+            convert_step_to_command,
+            create_bundle_for_flow_run,
+        )
+
+        if (
+            self.work_pool.storage_configuration.bundle_upload_step is None
+            or self.work_pool.storage_configuration.bundle_execution_step is None
+        ):
+            raise RuntimeError(
+                f"Storage is not configured for work pool {self.work_pool.name!r}. "
+                "Please configure storage for the work pool by running `prefect "
+                "work-pool storage configure`."
+            )
+
+        bundle_key = str(uuid.uuid4())
+        upload_command = convert_step_to_command(
+            self.work_pool.storage_configuration.bundle_upload_step,
+            bundle_key,
+            quiet=True,
+        )
+        execute_command = convert_step_to_command(
+            self.work_pool.storage_configuration.bundle_execution_step, bundle_key
+        )
+
+        job_variables = (job_variables or {}) | {"command": " ".join(execute_command)}
+        flow_run = await self.client.create_flow_run(
+            flow,
+            parameters=parameters,
+            state=Pending(),
+            job_variables=job_variables,
+            work_pool_name=self.work_pool.name,
+        )
+        if task_status is not None:
+            # Emit the flow run object to .submit to allow it to return a future as soon as possible
+            task_status.started(flow_run)
+        # Avoid an API call to get the flow
+        api_flow = APIFlow(id=flow_run.flow_id, name=flow.name, labels={})
+        logger = self.get_flow_run_logger(flow_run)
+
+        configuration = await self.job_configuration.from_template_and_values(
+            base_job_template=self.work_pool.base_job_template,
+            values=job_variables,
+            client=self._client,
+        )
+        configuration.prepare_for_flow_run(
+            flow_run=flow_run,
+            flow=api_flow,
+            work_pool=self.work_pool,
+            worker_name=self.name,
+        )
+
+        bundle = create_bundle_for_flow_run(flow=flow, flow_run=flow_run)
+
+        with tempfile.TemporaryDirectory() as temp_dir:
+            await (
+                anyio.Path(temp_dir)
+                .joinpath(bundle_key)
+                .write_bytes(json.dumps(bundle).encode("utf-8"))
+            )
+
+            try:
+                full_command = upload_command + [bundle_key]
+                logger.debug(
+                    "Uploading execution bundle with command: %s", full_command
+                )
+                await anyio.run_process(
+                    full_command,
+                    cwd=temp_dir,
+                )
+            except subprocess.CalledProcessError as e:
+                raise RuntimeError(e.stderr.decode("utf-8")) from e
+
+        logger.debug("Successfully uploaded execution bundle")
+
+        try:
+            result = await self.run(flow_run, configuration)
+
+            if result.status_code != 0:
+                await self._propose_crashed_state(
+                    flow_run,
+                    (
+                        "Flow run infrastructure exited with non-zero status code"
+                        f" {result.status_code}."
+                    ),
+                )
+        except Exception as exc:
+            # This flow run was being submitted and did not start successfully
+            logger.exception(
+                f"Failed to submit flow run '{flow_run.id}' to infrastructure."
+            )
+            message = f"Flow run could not be submitted to infrastructure:\n{exc!r}"
+            await self._propose_crashed_state(flow_run, message, client=self.client)
+
     @classmethod
     def __dispatch_key__(cls) -> str | None:
         if cls.__name__ == "BaseWorker":
@@ -1230,11 +1379,13 @@ class BaseWorker(abc.ABC, Generic[C, V, R]):
                 exc_info=True,
             )
 
-    async def _propose_crashed_state(self, flow_run: "FlowRun", message: str) -> None:
+    async def _propose_crashed_state(
+        self, flow_run: "FlowRun", message: str, client: PrefectClient | None = None
+    ) -> None:
         run_logger = self.get_flow_run_logger(flow_run)
         try:
             state = await propose_state(
-                self.client,
+                client or self.client,
                 Crashed(message=message),
                 flow_run_id=flow_run.id,
             )
@@ -1345,8 +1496,16 @@ class BaseWorker(abc.ABC, Generic[C, V, R]):
         return self
 
     async def __aexit__(self, *exc_info: Any) -> None:
-        self._logger.debug("Exiting worker context...")
-        await self.teardown(*exc_info)
+        try:
+            self._logger.debug("Exiting worker context...")
+            await self.teardown(*exc_info)
+        except (ExceptionGroup, BaseExceptionGroup) as exc:
+            # For less verbose tracebacks
+            exceptions = exc.exceptions
+            if len(exceptions) == 1:
+                raise exceptions[0] from None
+            else:
+                raise
 
     def __repr__(self) -> str:
         return f"Worker(pool={self._work_pool_name!r}, name={self.name!r})"