prefect-client 3.6.23.dev3__tar.gz → 3.6.24.dev1__tar.gz

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prefect-client
-Version: 3.6.23.dev3
+Version: 3.6.24.dev1
 Summary: Workflow orchestration and management.
 Project-URL: Changelog, https://github.com/PrefectHQ/prefect/releases
 Project-URL: Documentation, https://docs.prefect.io

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/AGENTS.md

@@ -46,4 +46,7 @@ Use `get_logger()` from `prefect.logging` instead of raw `logging.getLogger()`
 - `settings/` → Configuration system (see settings/AGENTS.md)
 - `concurrency/` → Concurrency slot acquisition and lease management (see concurrency/AGENTS.md)
 - `logging/` → Logging handlers, API log shipping, and run-context loggers (see logging/AGENTS.md)
+- `runner/` → Thin facade over extracted single-responsibility classes for local flow run execution (see runner/AGENTS.md)
+- `deployments/` → YAML-driven deployment lifecycle: project init, build/push/pull steps, and triggering remote flow runs (see deployments/AGENTS.md)
+- `docker/` → `DockerImage` class for building and pushing Docker images during deployment
 - `telemetry/` → OS-level resource metric collection and run telemetry

prefect_client-3.6.24.dev1/src/prefect/_build_info.py

@@ -0,0 +1,5 @@
+# Generated by versioningit
+__version__ = "3.6.24.dev1"
+__build_date__ = "2026-03-21 08:11:26.248833+00:00"
+__git_commit__ = "a74453a3544b6e146a7b422d9bfb3d9bd0a55594"
+__dirty__ = False

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/blocks/core.py

@@ -609,7 +609,7 @@ class Block(BaseModel, ABC):
         # reported from https://github.com/PrefectHQ/prefect-dbt/issues/54
         data_keys = self.model_json_schema(by_alias=False)["properties"].keys()
 
-        # `block_document_data`` must return the aliased version for it to show in the UI
+        # `block_document_data` must return the aliased version for it to show in the UI
         block_document_data = self.model_dump(
             by_alias=True,
             include=data_keys,

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/_version_checking.py

@@ -81,20 +81,20 @@ async def check_server_version(
     """Perform a one-shot server version compatibility check.
 
     The check is skipped when:
-    * The ``server_version_check_enabled`` setting is ``False``.
+    * The `server_version_check_enabled` setting is `False`.
     * The *api_url* points at Prefect Cloud (Cloud is always compatible).
     * A check for this *(api_url, client_version)* pair has already passed.
 
-    On a major-version mismatch a ``RuntimeError`` is raised (regardless of
+    On a major-version mismatch a `RuntimeError` is raised (regardless of
     *raise_on_error*).  When the server is simply older than the client a
     warning is logged.
 
     Args:
-        api_url: The base Prefect API URL (e.g. ``http://localhost:4200/api``).
+        api_url: The base Prefect API URL (e.g. `http://localhost:4200/api`).
         logger: Logger used for warnings and debug messages.
-        raise_on_error: When ``True`` (the default, used by HTTP clients),
-            raise ``RuntimeError`` if the version endpoint cannot be reached.
-            When ``False`` (used by WebSocket clients), log a debug message
+        raise_on_error: When `True` (the default, used by HTTP clients),
+            raise `RuntimeError` if the version endpoint cannot be reached.
+            When `False` (used by WebSocket clients), log a debug message
             and return silently so that the caller can still attempt its
             connection.
     """

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/orchestration/__init__.py

@@ -143,6 +143,7 @@ from prefect.client._version_checking import (
 )
 
 if TYPE_CHECKING:
+    from prefect.client.schemas.responses import WorkQueueConcurrencyStatus
     from prefect.tasks import Task as TaskObject
 
 from prefect.client.base import (
@@ -713,6 +714,44 @@ class PrefectClient(
                 raise
         return WorkQueueStatusDetail.model_validate(response.json())
 
+    async def read_work_queue_concurrency_status(
+        self,
+        id: UUID,
+        page: int = 1,
+        limit: Optional[int] = None,
+    ) -> "WorkQueueConcurrencyStatus":
+        """
+        Read concurrency status for a work queue.
+
+        Args:
+            id: the id of the work queue
+            page: Page number (1-indexed).
+            limit: Max flow runs per page (server default if None).
+
+        Raises:
+            prefect.exceptions.ObjectNotFound: If request returns 404
+            httpx.RequestError: If request fails
+
+        Returns:
+            Paginated WorkQueueConcurrencyStatus with flow run summaries
+        """
+        from prefect.client.schemas.responses import WorkQueueConcurrencyStatus
+
+        body: dict = {"page": page}
+        if limit is not None:
+            body["limit"] = limit
+
+        try:
+            response = await self._client.post(
+                f"/work_queues/{id}/concurrency_status", json=body
+            )
+        except httpx.HTTPStatusError as e:
+            if e.response.status_code == status.HTTP_404_NOT_FOUND:
+                raise prefect.exceptions.ObjectNotFound(http_exc=e) from e
+            else:
+                raise
+        return WorkQueueConcurrencyStatus.model_validate(response.json())
+
     async def match_work_queues(
         self,
         prefixes: list[str],
@@ -1882,6 +1921,44 @@ class SyncPrefectClient(
                 raise
         return WorkQueueStatusDetail.model_validate(response.json())
 
+    def read_work_queue_concurrency_status(
+        self,
+        id: UUID,
+        page: int = 1,
+        limit: Optional[int] = None,
+    ) -> "WorkQueueConcurrencyStatus":
+        """
+        Read concurrency status for a work queue.
+
+        Args:
+            id: the id of the work queue
+            page: Page number (1-indexed).
+            limit: Max flow runs per page (server default if None).
+
+        Raises:
+            prefect.exceptions.ObjectNotFound: If request returns 404
+            httpx.RequestError: If request fails
+
+        Returns:
+            Paginated WorkQueueConcurrencyStatus with flow run summaries
+        """
+        from prefect.client.schemas.responses import WorkQueueConcurrencyStatus
+
+        body: dict = {"page": page}
+        if limit is not None:
+            body["limit"] = limit
+
+        try:
+            response = self._client.post(
+                f"/work_queues/{id}/concurrency_status", json=body
+            )
+        except httpx.HTTPStatusError as e:
+            if e.response.status_code == status.HTTP_404_NOT_FOUND:
+                raise prefect.exceptions.ObjectNotFound(http_exc=e) from e
+            else:
+                raise
+        return WorkQueueConcurrencyStatus.model_validate(response.json())
+
     def match_work_queues(
         self,
         prefixes: list[str],

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/orchestration/_work_pools/client.py

@@ -25,7 +25,10 @@ if TYPE_CHECKING:
         WorkerMetadata,
         WorkPool,
     )
-    from prefect.client.schemas.responses import WorkerFlowRunResponse
+    from prefect.client.schemas.responses import (
+        WorkerFlowRunResponse,
+        WorkPoolConcurrencyStatus,
+    )
 
 from prefect.exceptions import ObjectAlreadyExists, ObjectNotFound, ObjectUnsupported
 
@@ -316,6 +319,47 @@ class WorkPoolClient(BaseClient):
 
         return WorkerFlowRunResponse.model_validate_list(response.json())
 
+    def read_work_pool_concurrency_status(
+        self,
+        work_pool_name: str,
+        page: int = 1,
+        limit: int | None = None,
+        flow_run_limit: int = 10,
+    ) -> "WorkPoolConcurrencyStatus":
+        """
+        Reads concurrency status for a work pool.
+
+        Args:
+            work_pool_name: The name of the work pool.
+            page: Page number (1-indexed).
+            limit: Max queues per page (server default if None).
+            flow_run_limit: Max flow runs per queue (0-200).
+
+        Returns:
+            Paginated concurrency status with per-queue breakdown.
+        """
+        from prefect.client.schemas.responses import WorkPoolConcurrencyStatus
+
+        body: dict[str, Any] = {"page": page}
+        if limit is not None:
+            body["limit"] = limit
+        body["flow_run_limit"] = flow_run_limit
+
+        try:
+            response = self.request(
+                "POST",
+                "/work_pools/{name}/concurrency_status",
+                path_params={"name": work_pool_name},
+                json=body,
+            )
+        except HTTPStatusError as e:
+            if e.response.status_code == 404:
+                raise ObjectNotFound(http_exc=e) from e
+            else:
+                raise
+
+        return WorkPoolConcurrencyStatus.model_validate(response.json())
+
 
 class WorkPoolAsyncClient(BaseAsyncClient):
     async def send_worker_heartbeat(
@@ -599,3 +643,44 @@ class WorkPoolAsyncClient(BaseAsyncClient):
                 raise
 
         return WorkerFlowRunResponse.model_validate_list(response.json())
+
+    async def read_work_pool_concurrency_status(
+        self,
+        work_pool_name: str,
+        page: int = 1,
+        limit: int | None = None,
+        flow_run_limit: int = 10,
+    ) -> "WorkPoolConcurrencyStatus":
+        """
+        Reads concurrency status for a work pool.
+
+        Args:
+            work_pool_name: The name of the work pool.
+            page: Page number (1-indexed).
+            limit: Max queues per page (server default if None).
+            flow_run_limit: Max flow runs per queue (0-200).
+
+        Returns:
+            Paginated concurrency status with per-queue breakdown.
+        """
+        from prefect.client.schemas.responses import WorkPoolConcurrencyStatus
+
+        body: dict[str, Any] = {"page": page}
+        if limit is not None:
+            body["limit"] = limit
+        body["flow_run_limit"] = flow_run_limit
+
+        try:
+            response = await self.request(
+                "POST",
+                "/work_pools/{name}/concurrency_status",
+                path_params={"name": work_pool_name},
+                json=body,
+            )
+        except HTTPStatusError as e:
+            if e.response.status_code == 404:
+                raise ObjectNotFound(http_exc=e) from e
+            else:
+                raise
+
+        return WorkPoolConcurrencyStatus.model_validate(response.json())

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/orchestration/routes.py

@@ -125,6 +125,7 @@ ServerRoutes = Literal[
     "/version",
     "/work_pools/",
     "/work_pools/{name}",
+    "/work_pools/{name}/concurrency_status",
     "/work_pools/{name}/get_scheduled_flow_runs",
     "/work_pools/{work_pool_name}/queues",
     "/work_pools/{work_pool_name}/queues/{name}",
@@ -137,6 +138,7 @@ ServerRoutes = Literal[
     "/work_queues/",
     "/work_queues/{id}",
     "/work_queues/{id}/get_runs",
+    "/work_queues/{id}/concurrency_status",
     "/work_queues/{id}/status",
     "/work_queues/filter",
     "/work_queues/name/{name}",

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/schemas/objects.py

@@ -1433,6 +1433,13 @@ class WorkQueue(ObjectBaseModel):
     status: Optional[WorkQueueStatus] = Field(
         default=None, description="The queue status."
     )
+    active_slots: Optional[int] = Field(
+        default=None,
+        description=(
+            "The number of concurrency slots currently in use. "
+            "None when concurrency_limit is not set."
+        ),
+    )
 
 
 class WorkQueueHealthPolicy(PrefectBaseModel):
@@ -1554,6 +1561,13 @@ class WorkPool(ObjectBaseModel):
     status: Optional[WorkPoolStatus] = Field(
         default=None, description="The current status of the work pool."
     )
+    active_slots: Optional[int] = Field(
+        default=None,
+        description=(
+            "The number of concurrency slots occupied by pending or running "
+            "flow runs. None when concurrency_limit is not set."
+        ),
+    )
 
     storage_configuration: WorkPoolStorageConfiguration = Field(
         default_factory=WorkPoolStorageConfiguration,

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/client/schemas/responses.py

@@ -1,8 +1,18 @@
 import datetime
-from typing import TYPE_CHECKING, Any, ClassVar, Generic, Optional, TypeVar, Union
+from datetime import timedelta
+from typing import (
+    TYPE_CHECKING,
+    Annotated,
+    Any,
+    ClassVar,
+    Generic,
+    Optional,
+    TypeVar,
+    Union,
+)
 from uuid import UUID
 
-from pydantic import ConfigDict, Field
+from pydantic import BeforeValidator, ConfigDict, Field
 from typing_extensions import Literal
 
 import prefect.client.schemas.objects as objects
@@ -501,3 +511,55 @@ class GlobalConcurrencyLimitResponse(ObjectBaseModel):
         default=2.0,
         description="The decay rate for active slots when used as a rate limit.",
     )
+
+
+class FlowRunSlotSummary(PrefectBaseModel):
+    """Summary of a flow run occupying a concurrency slot."""
+
+    id: UUID
+    name: str
+    state_type: Optional[objects.StateType] = None
+    state_name: Optional[str] = None
+    start_time: Optional[DateTime] = None
+    state_timestamp: Optional[DateTime] = None
+    time_in_current_state: Optional[timedelta] = None
+
+
+def _coerce_active_slots(v: int | None) -> int:
+    """Cloud may return null for active_slots; coerce to 0."""
+    return v if v is not None else 0
+
+
+class WorkQueueConcurrencyStatusDetail(PrefectBaseModel):
+    """Per-queue concurrency status with flow run details."""
+
+    queue_id: UUID
+    queue_name: str
+    active_slots: Annotated[int, BeforeValidator(_coerce_active_slots)]
+    concurrency_limit: Optional[int] = None
+    flow_runs: list[FlowRunSlotSummary] = Field(default_factory=list)
+    flow_run_count: Optional[int] = None
+
+
+class WorkPoolConcurrencyStatus(PrefectBaseModel):
+    """Paginated pool-level concurrency status with per-queue breakdown."""
+
+    active_slots: Annotated[int, BeforeValidator(_coerce_active_slots)]
+    concurrency_limit: Optional[int] = None
+    queues: list[WorkQueueConcurrencyStatusDetail] = Field(default_factory=list)
+    count: int
+    limit: int
+    pages: int
+    page: int
+
+
+class WorkQueueConcurrencyStatus(PrefectBaseModel):
+    """Paginated queue-level concurrency status with flow run details."""
+
+    active_slots: Annotated[int, BeforeValidator(_coerce_active_slots)]
+    concurrency_limit: Optional[int] = None
+    flow_runs: list[FlowRunSlotSummary] = Field(default_factory=list)
+    count: int
+    limit: int
+    pages: int
+    page: int

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/concurrency/AGENTS.md

@@ -38,5 +38,5 @@ Layered — public → internal → services, with leases and events as cross-cu
 
 - **Service singleton is keyed on `frozenset(names)`.** Passing the same names in different order reuses the same singleton; passing a strict subset creates a different singleton. Each unique name-set gets its own queue.
 - **Lease renewal runs on the global event loop** (sync path). If the global loop is blocked or torn down, renewal silently fails — you'll see the renewal failure callback fire after `max_attempts` retries are exhausted.
-- **Cancellation during acquire** — lease IDs for mid-acquisition slots land in `ConcurrencyContext.cleanup_lease_ids`. `ConcurrencyContext.__exit__` releases them via a sync client. If `ConcurrencyContext` is not active, those leases are abandoned until server-side expiry.
+- **Cancellation during acquire or release** — if a `CancelledError` is raised either during slot acquisition or during `release_concurrency_slots_with_lease` in the `finally` block, the lease ID is appended to `ConcurrencyContext.cleanup_lease_ids`. `ConcurrencyContext.__exit__` releases them via a sync client. If `ConcurrencyContext` is not active, those leases are abandoned until server-side expiry.
 - **`v1/` subdirectory** contains the legacy slot API (no lease model). New code should use the top-level `concurrency()` / `rate_limit()` APIs.

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/concurrency/_sync.py

@@ -4,6 +4,7 @@ from contextlib import contextmanager
 from typing import Generator, Literal, Optional
 from uuid import UUID
 
+from prefect._internal.concurrency.cancellation import CancelledError
 from prefect.client.schemas.objects import ConcurrencyLeaseHolder
 from prefect.client.schemas.responses import (
     ConcurrencyLimitWithLeaseResponse,
@@ -19,6 +20,7 @@ from prefect.concurrency._events import (
     emit_concurrency_release_events,
 )
 from prefect.concurrency._leases import maintain_concurrency_lease
+from prefect.concurrency.context import ConcurrencyContext
 from prefect.utilities.asyncutils import run_coro_as_sync
 
 
@@ -146,7 +148,15 @@ def concurrency(
         ):
             yield
     finally:
-        release_concurrency_slots_with_lease(acquisition_response.lease_id)
+        try:
+            release_concurrency_slots_with_lease(acquisition_response.lease_id)
+        except CancelledError:
+            # The task was cancelled before it could release the lease. Add the
+            # lease ID to the cleanup list so it can be released when the
+            # concurrency context is exited.
+            if ctx := ConcurrencyContext.get():
+                ctx.cleanup_lease_ids.append(acquisition_response.lease_id)
+
         emit_concurrency_release_events(
             acquisition_response.limits, occupy, emitted_events
         )

prefect_client-3.6.24.dev1/src/prefect/deployments/AGENTS.md

@@ -0,0 +1,39 @@
+# Deployments
+
+YAML-driven configuration for packaging, publishing, and triggering flow runs from infrastructure.
+
+## Purpose & Scope
+
+Handles deployment lifecycle: initializing projects, building/pushing deployment artifacts, and triggering remote flow runs. Does NOT manage flow execution itself — that lives in `flow_engine.py` and `task_engine.py`.
+
+## Entry Points & Contracts
+
+- `runner.py` → `deploy()` — programmatic deployment creation from `Flow` objects
+- `flow_runs.py` → `run_deployment()` / `arun_deployment()` — trigger a run of an existing deployment
+- `base.py` → `initialize_project()` — scaffold `prefect.yaml` in a project directory
+- `steps/core.py` → `run_step()` / `run_steps()` — execute lifecycle steps defined in `prefect.yaml`
+
+## Steps System
+
+Steps are YAML entries in `build`, `push`, or `pull` blocks of `prefect.yaml`. Each step maps to a Python function imported at runtime. The `requires` keyword auto-installs missing packages before import.
+
+Step outputs are templated into subsequent steps via `{{ step-id.key }}`.
+
+Built-in steps:
+- `steps/pull.py` — `git_clone`, `set_working_directory`, `pull_from_remote_storage`
+- `steps/utility.py` — `run_shell_script`, `pip_install_requirements`
+
+## Entrypoint Formats
+
+`runner.py`'s `from_storage` / `afrom_storage` (and `Flow.from_source`) support two entrypoint formats:
+- **File path**: `path/to/file.py:flow_func_name` — detected by presence of `:`
+- **Module path**: `my_package.flows.flow_func` — detected by absence of `:`
+
+For module path entrypoints, the storage destination is temporarily prepended to `sys.path` so the module can be imported, then removed in a `finally` block. Any new code that loads flows from module paths must follow this same pattern to avoid polluting `sys.path`.
+
+## Pitfalls
+
+- **Windows shell mode**: `run_shell_script` always uses `asyncio.create_subprocess_shell` on Windows (`sys.platform == "win32"`), regardless of the `shell` parameter. This ensures cmd.exe built-ins (`echo`, `dir`, `set`, etc.) work. On non-Windows, `shell=False` (default) uses `create_subprocess_exec` with `shlex.split`.
+- **Step ID namespace**: `id` and `requires` are reserved keywords — do not use them as step output keys.
+- **Step import side effects**: steps are imported dynamically; packages listed in `requires` are installed into the current environment at execution time.
+- **String `image` argument suppresses build/push output.** When `deploy()` / `adeploy()` receives `image` as a plain string, it constructs `DockerImage(stream_progress_to=None)`, silencing all build and push progress. `DockerImage` itself defaults to `sys.stdout`, so users who pass a string get no output. To see build/push progress, pass a `DockerImage` object explicitly: `DockerImage("registry/image:tag", stream_progress_to=sys.stdout)`.

{prefect_client-3.6.23.dev3 → prefect_client-3.6.24.dev1}/src/prefect/deployments/runner.py

@@ -32,6 +32,7 @@ Example:
 from __future__ import annotations
 
 import importlib
+import sys
 import tempfile
 from datetime import datetime, timedelta
 from pathlib import Path
@@ -1160,7 +1161,8 @@ class RunnerDeployment(BaseModel):
 
         Args:
             entrypoint:  The path to a file containing a flow and the name of the flow function in
-                the format `./path/to/file.py:flow_func_name`.
+                the format `./path/to/file.py:flow_func_name`, or a module path to a flow function
+                in the format `module.path.flow_func_name`.
             name: A name for the deployment
             flow_name: The name of the flow to deploy
             storage: A storage object to use for retrieving flow code. If not provided, a
@@ -1213,10 +1215,19 @@ class RunnerDeployment(BaseModel):
             storage.set_base_path(Path(tmpdir))
             await storage.pull_code()
 
-            full_entrypoint = str(storage.destination / entrypoint)
-            flow = await from_async.wait_for_call_in_new_thread(
-                create_call(load_flow_from_entrypoint, full_entrypoint)
-            )
+            if ":" in entrypoint:
+                full_entrypoint = str(storage.destination / entrypoint)
+            else:
+                sys.path.insert(0, str(storage.destination))
+                full_entrypoint = entrypoint
+
+            try:
+                flow = await from_async.wait_for_call_in_new_thread(
+                    create_call(load_flow_from_entrypoint, full_entrypoint)
+                )
+            finally:
+                if ":" not in entrypoint:
+                    sys.path.remove(str(storage.destination))
 
         deployment = cls(
             name=name,
@@ -1239,6 +1250,11 @@ class RunnerDeployment(BaseModel):
             job_variables=job_variables,
         )
         deployment._sla = _sla
+        deployment._entrypoint_type = (
+            EntrypointType.FILE_PATH
+            if ":" in entrypoint
+            else EntrypointType.MODULE_PATH
+        )
         deployment._path = str(storage.destination).replace(
             tmpdir, "$STORAGE_BASE_PATH"
         )
@@ -1282,7 +1298,8 @@ class RunnerDeployment(BaseModel):
 
         Args:
             entrypoint:  The path to a file containing a flow and the name of the flow function in
-                the format `./path/to/file.py:flow_func_name`.
+                the format `./path/to/file.py:flow_func_name`, or a module path to a flow function
+                in the format `module.path.flow_func_name`.
             name: A name for the deployment
             flow_name: The name of the flow to deploy
             storage: A storage object to use for retrieving flow code. If not provided, a
@@ -1335,8 +1352,17 @@ class RunnerDeployment(BaseModel):
             storage.set_base_path(Path(tmpdir))
             run_coro_as_sync(storage.pull_code())
 
-            full_entrypoint = str(storage.destination / entrypoint)
-            flow = load_flow_from_entrypoint(full_entrypoint)
+            if ":" in entrypoint:
+                full_entrypoint = str(storage.destination / entrypoint)
+            else:
+                sys.path.insert(0, str(storage.destination))
+                full_entrypoint = entrypoint
+
+            try:
+                flow = load_flow_from_entrypoint(full_entrypoint)
+            finally:
+                if ":" not in entrypoint:
+                    sys.path.remove(str(storage.destination))
 
         deployment = cls(
             name=name,
@@ -1359,6 +1385,11 @@ class RunnerDeployment(BaseModel):
             job_variables=job_variables,
         )
         deployment._sla = _sla
+        deployment._entrypoint_type = (
+            EntrypointType.FILE_PATH
+            if ":" in entrypoint
+            else EntrypointType.MODULE_PATH
+        )
         deployment._path = str(storage.destination).replace(
             tmpdir, "$STORAGE_BASE_PATH"
         )
@@ -1448,7 +1479,7 @@ async def adeploy(
 
     if image and isinstance(image, str):
         image_name, image_tag = parse_image_tag(image)
-        image = DockerImage(name=image_name, tag=image_tag)
+        image = DockerImage(name=image_name, tag=image_tag, stream_progress_to=None)
 
     try:
         async with get_client() as client:
@@ -1687,7 +1718,7 @@ def deploy(
 
     if image and isinstance(image, str):
         image_name, image_tag = parse_image_tag(image)
-        image = DockerImage(name=image_name, tag=image_tag)
+        image = DockerImage(name=image_name, tag=image_tag, stream_progress_to=None)
 
     try:
         with get_client(sync_client=True) as client: