modal 0.73.74__py3-none-any.whl → 0.73.76__py3-none-any.whl

modal/_functions.py

@@ -417,18 +417,18 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
         allow_cross_region_volumes: bool = False,
         volumes: dict[Union[str, PurePosixPath], Union[_Volume, _CloudBucketMount]] = {},
         webhook_config: Optional[api_pb2.WebhookConfig] = None,
+        cpu: Optional[Union[float, tuple[float, float]]] = None,
         memory: Optional[Union[int, tuple[int, int]]] = None,
         proxy: Optional[_Proxy] = None,
         retries: Optional[Union[int, Retries]] = None,
         timeout: Optional[int] = None,
-        concurrency_limit: Optional[int] = None,
+        min_containers: Optional[int] = None,
+        max_containers: Optional[int] = None,
+        buffer_containers: Optional[int] = None,
+        scaledown_window: Optional[int] = None,
         allow_concurrent_inputs: Optional[int] = None,
         batch_max_size: Optional[int] = None,
         batch_wait_ms: Optional[int] = None,
-        container_idle_timeout: Optional[int] = None,
-        cpu: Optional[Union[float, tuple[float, float]]] = None,
-        # keep_warm=True is equivalent to keep_warm=1
-        keep_warm: Optional[int] = None,
         cloud: Optional[str] = None,
         scheduler_placement: Optional[SchedulerPlacement] = None,
         is_builder_function: bool = False,
@@ -442,7 +442,6 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
         ephemeral_disk: Optional[int] = None,
         # current default: first-party, future default: main-package
         include_source: Optional[bool] = None,
-        _experimental_buffer_containers: Optional[int] = None,
         _experimental_proxy_ip: Optional[str] = None,
         _experimental_custom_scaling_factor: Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,
@@ -564,20 +563,21 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
                     force_build=image.force_build or pf.force_build,
                 )
 
-        if keep_warm is not None and not isinstance(keep_warm, int):
-            raise TypeError(f"`keep_warm` must be an int or bool, not {type(keep_warm).__name__}")
-
-        if (keep_warm is not None) and (concurrency_limit is not None) and concurrency_limit < keep_warm:
+        # Note that we also do these checks in FunctionCreate; could drop them here
+        if min_containers is not None and not isinstance(min_containers, int):
+            raise InvalidError(f"`min_containers` must be an int, not {type(min_containers).__name__}")
+        if min_containers is not None and max_containers is not None and max_containers < min_containers:
             raise InvalidError(
-                f"Function `{info.function_name}` has `{concurrency_limit=}`, "
-                f"strictly less than its `{keep_warm=}` parameter."
+                f"`min_containers` ({min_containers}) cannot be greater than `max_containers` ({max_containers})"
             )
+        if scaledown_window is not None and scaledown_window <= 0:
+            raise InvalidError("`scaledown_window` must be > 0")
 
         autoscaler_settings = api_pb2.AutoscalerSettings(
-            max_containers=concurrency_limit,
-            min_containers=keep_warm,
-            buffer_containers=_experimental_buffer_containers,
-            scaledown_window=container_idle_timeout,
+            min_containers=min_containers,
+            max_containers=max_containers,
+            buffer_containers=buffer_containers,
+            scaledown_window=scaledown_window,
         )
 
         if _experimental_custom_scaling_factor is not None and (
@@ -605,9 +605,6 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
                 if arg.default is not inspect.Parameter.empty:
                     raise InvalidError(f"Modal batched function {func_name} does not accept default arguments.")
 
-        if container_idle_timeout is not None and container_idle_timeout <= 0:
-            raise InvalidError("`container_idle_timeout` must be > 0")
-
         if max_inputs is not None:
             if not isinstance(max_inputs, int):
                 raise InvalidError(f"`max_inputs` must be an int, not {type(max_inputs).__name__}")
@@ -773,11 +770,8 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
                     proxy_id=(proxy.object_id if proxy else None),
                     retry_policy=retry_policy,
                     timeout_secs=timeout_secs or 0,
-                    task_idle_timeout_secs=container_idle_timeout or 0,
-                    concurrency_limit=concurrency_limit or 0,
                     pty_info=pty_info,
                     cloud_provider_str=cloud if cloud else "",
-                    warm_pool_size=keep_warm or 0,
                     runtime=config.get("function_runtime"),
                     runtime_debug=config.get("function_runtime_debug"),
                     runtime_perf_record=config.get("runtime_perf_record"),
@@ -802,10 +796,15 @@ class _Function(typing.Generic[P, ReturnType, OriginalReturnType], _Object, type
                     snapshot_debug=config.get("snapshot_debug"),
                     _experimental_group_size=cluster_size or 0,  # Experimental: Clustered functions
                     _experimental_concurrent_cancellations=True,
-                    _experimental_buffer_containers=_experimental_buffer_containers or 0,
                     _experimental_proxy_ip=_experimental_proxy_ip,
                     _experimental_custom_scaling=_experimental_custom_scaling_factor is not None,
                     _experimental_enable_gpu_snapshot=_experimental_enable_gpu_snapshot,
+                    # --- These are deprecated in favor of autoscaler_settings
+                    warm_pool_size=min_containers or 0,
+                    concurrency_limit=max_containers or 0,
+                    _experimental_buffer_containers=buffer_containers or 0,
+                    task_idle_timeout_secs=scaledown_window or 0,
+                    # ---
                 )
 
                 if isinstance(gpu, list):

modal/_utils/deprecation.py

@@ -87,3 +87,38 @@ def renamed_parameter(
         return wrapper
 
     return decorator
+
+
+def warn_on_renamed_autoscaler_settings(func: Callable[P, R]) -> Callable[P, R]:
+    name_map = {
+        "keep_warm": "min_containers",
+        "concurrency_limit": "max_containers",
+        "_experimental_buffer_containers": "buffer_containers",
+        "container_idle_timeout": "scaledown_window",
+    }
+
+    @functools.wraps(func)
+    def wrapper(*args: P.args, **kwargs: P.kwargs) -> R:
+        mut_kwargs: dict[str, Any] = locals()["kwargs"]  # Avoid referencing kwargs directly due to bug in sigtools
+
+        substitutions = []
+        old_params_used = name_map.keys() & mut_kwargs.keys()
+        for old_param, new_param in name_map.items():
+            if old_param in old_params_used:
+                new_param = name_map[old_param]
+                mut_kwargs[new_param] = mut_kwargs.pop(old_param)
+                substitutions.append(f"- {old_param} -> {new_param}")
+
+        if substitutions:
+            substitution_string = "\n".join(substitutions)
+            message = (
+                "We have renamed several parameters related to autoscaling."
+                " Please update your code to use the following new names:"
+                f"\n\n{substitution_string}"
+                "\n\nSee https://modal.com/docs/guide/modal-1-0-migration for more details."
+            )
+            deprecation_warning((2025, 2, 24), message, pending=True, show_source=True)
+
+        return func(*args, **kwargs)
+
+    return wrapper

modal/app.py

@@ -29,7 +29,12 @@ from ._partial_function import (
     _PartialFunctionFlags,
 )
 from ._utils.async_utils import synchronize_api
-from ._utils.deprecation import deprecation_error, deprecation_warning, renamed_parameter
+from ._utils.deprecation import (
+    deprecation_error,
+    deprecation_warning,
+    renamed_parameter,
+    warn_on_renamed_autoscaler_settings,
+)
 from ._utils.function_utils import FunctionInfo, is_global_object, is_method_fn
 from ._utils.grpc_utils import retry_transient_errors
 from ._utils.mount_utils import validate_volumes
@@ -559,6 +564,7 @@ class _App:
 
         return wrapped
 
+    @warn_on_renamed_autoscaler_settings
     def function(
         self,
         _warn_parentheses_missing: Any = None,
@@ -586,17 +592,14 @@ class _App:
         # Or, pass (request, limit) to additionally specify a hard limit in MiB.
         memory: Optional[Union[int, tuple[int, int]]] = None,
         ephemeral_disk: Optional[int] = None,  # Specify, in MiB, the ephemeral disk size for the Function.
+        min_containers: Optional[int] = None,  # Minimum number of containers to keep warm, even when Function is idle.
+        max_containers: Optional[int] = None,  # Limit on the number of containers that can be concurrently running.
+        buffer_containers: Optional[int] = None,  # Number of additional idle containers to maintain under active load.
+        scaledown_window: Optional[int] = None,  # Max amount of time a container can remain idle before scaling down.
         proxy: Optional[_Proxy] = None,  # Reference to a Modal Proxy to use in front of this function.
         retries: Optional[Union[int, Retries]] = None,  # Number of times to retry each input in case of failure.
-        concurrency_limit: Optional[
-            int
-        ] = None,  # An optional maximum number of concurrent containers running the function (keep_warm sets minimum).
         allow_concurrent_inputs: Optional[int] = None,  # Number of inputs the container may fetch to run concurrently.
-        container_idle_timeout: Optional[int] = None,  # Timeout for idle containers waiting for inputs to shut down.
         timeout: Optional[int] = None,  # Maximum execution time of the function in seconds.
-        keep_warm: Optional[
-            int
-        ] = None,  # An optional minimum number of containers to always keep warm (use concurrency_limit for maximum).
         name: Optional[str] = None,  # Sets the Modal name of the function within the app
         is_generator: Optional[
             bool
@@ -615,10 +618,14 @@ class _App:
         _experimental_scheduler_placement: Optional[
             SchedulerPlacement
         ] = None,  # Experimental controls over fine-grained scheduling (alpha).
-        _experimental_buffer_containers: Optional[int] = None,  # Number of additional, idle containers to keep around.
         _experimental_proxy_ip: Optional[str] = None,  # IP address of proxy
         _experimental_custom_scaling_factor: Optional[float] = None,  # Custom scaling factor
         _experimental_enable_gpu_snapshot: bool = False,  # Experimentally enable GPU memory snapshots.
+        # Parameters below here are deprecated. Please update your code as suggested
+        keep_warm: Optional[int] = None,  # Replaced with `min_containers`
+        concurrency_limit: Optional[int] = None,  # Replaced with `max_containers`
+        container_idle_timeout: Optional[int] = None,  # Replaced with `scaledown_window`
+        _experimental_buffer_containers: Optional[int] = None,  # Now stable API with `buffer_containers`
     ) -> _FunctionDecoratorType:
         """Decorator to register a new Modal [Function](/docs/reference/modal.Function) with this App."""
         if isinstance(_warn_parentheses_missing, _Image):
@@ -635,7 +642,7 @@ class _App:
         def wrapped(
             f: Union[_PartialFunction, Callable[..., Any], None],
         ) -> _Function:
-            nonlocal keep_warm, is_generator, cloud, serialized
+            nonlocal is_generator, cloud, serialized
 
             # Check if the decorated object is a class
             if inspect.isclass(f):
@@ -669,7 +676,6 @@ class _App:
                 raw_f = f.raw_f
                 webhook_config = f.webhook_config
                 is_generator = f.is_generator
-                keep_warm = f.keep_warm or keep_warm
                 batch_max_size = f.batch_max_size
                 batch_wait_ms = f.batch_wait_ms
             else:
@@ -743,20 +749,20 @@ class _App:
                 ephemeral_disk=ephemeral_disk,
                 proxy=proxy,
                 retries=retries,
-                concurrency_limit=concurrency_limit,
+                min_containers=min_containers,
+                max_containers=max_containers,
+                buffer_containers=buffer_containers,
+                scaledown_window=scaledown_window,
                 allow_concurrent_inputs=allow_concurrent_inputs,
                 batch_max_size=batch_max_size,
                 batch_wait_ms=batch_wait_ms,
-                container_idle_timeout=container_idle_timeout,
                 timeout=timeout,
-                keep_warm=keep_warm,
                 cloud=cloud,
                 webhook_config=webhook_config,
                 enable_memory_snapshot=enable_memory_snapshot,
                 block_network=block_network,
                 max_inputs=max_inputs,
                 scheduler_placement=scheduler_placement,
-                _experimental_buffer_containers=_experimental_buffer_containers,
                 _experimental_proxy_ip=_experimental_proxy_ip,
                 i6pn_enabled=i6pn_enabled,
                 cluster_size=cluster_size,  # Experimental: Clustered functions
@@ -771,6 +777,7 @@ class _App:
         return wrapped
 
     @typing_extensions.dataclass_transform(field_specifiers=(parameter,), kw_only_default=True)
+    @warn_on_renamed_autoscaler_settings
     def cls(
         self,
         _warn_parentheses_missing: Optional[bool] = None,
@@ -797,13 +804,14 @@ class _App:
         # Or, pass (request, limit) to additionally specify a hard limit in MiB.
         memory: Optional[Union[int, tuple[int, int]]] = None,
         ephemeral_disk: Optional[int] = None,  # Specify, in MiB, the ephemeral disk size for the Function.
+        min_containers: Optional[int] = None,  # Minimum number of containers to keep warm, even when Function is idle.
+        max_containers: Optional[int] = None,  # Limit on the number of containers that can be concurrently running.
+        buffer_containers: Optional[int] = None,  # Number of additional idle containers to maintain under active load.
+        scaledown_window: Optional[int] = None,  # Max amount of time a container can remain idle before scaling down.
         proxy: Optional[_Proxy] = None,  # Reference to a Modal Proxy to use in front of this function.
         retries: Optional[Union[int, Retries]] = None,  # Number of times to retry each input in case of failure.
-        concurrency_limit: Optional[int] = None,  # Limit for max concurrent containers running the function.
         allow_concurrent_inputs: Optional[int] = None,  # Number of inputs the container may fetch to run concurrently.
-        container_idle_timeout: Optional[int] = None,  # Timeout for idle containers waiting for inputs to shut down.
         timeout: Optional[int] = None,  # Maximum execution time of the function in seconds.
-        keep_warm: Optional[int] = None,  # An optional number of containers to always keep warm.
         cloud: Optional[str] = None,  # Cloud provider to run the function on. Possible values are aws, gcp, oci, auto.
         region: Optional[Union[str, Sequence[str]]] = None,  # Region or regions to run the function on.
         enable_memory_snapshot: bool = False,  # Enable memory checkpointing for faster cold starts.
@@ -816,10 +824,14 @@ class _App:
         _experimental_scheduler_placement: Optional[
             SchedulerPlacement
         ] = None,  # Experimental controls over fine-grained scheduling (alpha).
-        _experimental_buffer_containers: Optional[int] = None,  # Number of additional, idle containers to keep around.
         _experimental_proxy_ip: Optional[str] = None,  # IP address of proxy
         _experimental_custom_scaling_factor: Optional[float] = None,  # Custom scaling factor
         _experimental_enable_gpu_snapshot: bool = False,  # Experimentally enable GPU memory snapshots.
+        # Parameters below here are deprecated. Please update your code as suggested
+        keep_warm: Optional[int] = None,  # Replaced with `min_containers`
+        concurrency_limit: Optional[int] = None,  # Replaced with `max_containers`
+        container_idle_timeout: Optional[int] = None,  # Replaced with `scaledown_window`
+        _experimental_buffer_containers: Optional[int] = None,  # Now stable API with `buffer_containers`
     ) -> Callable[[CLS_T], CLS_T]:
         """
         Decorator to register a new Modal [Cls](/docs/reference/modal.Cls) with this App.
@@ -834,8 +846,6 @@ class _App:
             scheduler_placement = SchedulerPlacement(region=region)
 
         def wrapper(user_cls: CLS_T) -> CLS_T:
-            nonlocal keep_warm
-
             # Check if the decorated object is a class
             if not inspect.isclass(user_cls):
                 raise TypeError("The @app.cls decorator must be used on a class.")
@@ -873,25 +883,25 @@ class _App:
                 network_file_systems=network_file_systems,
                 allow_cross_region_volumes=allow_cross_region_volumes,
                 volumes={**self._volumes, **volumes},
+                cpu=cpu,
                 memory=memory,
                 ephemeral_disk=ephemeral_disk,
+                min_containers=min_containers,
+                max_containers=max_containers,
+                buffer_containers=buffer_containers,
+                scaledown_window=scaledown_window,
                 proxy=proxy,
                 retries=retries,
-                concurrency_limit=concurrency_limit,
                 allow_concurrent_inputs=allow_concurrent_inputs,
                 batch_max_size=batch_max_size,
                 batch_wait_ms=batch_wait_ms,
-                container_idle_timeout=container_idle_timeout,
                 timeout=timeout,
-                cpu=cpu,
-                keep_warm=keep_warm,
                 cloud=cloud,
                 enable_memory_snapshot=enable_memory_snapshot,
                 block_network=block_network,
                 max_inputs=max_inputs,
                 scheduler_placement=scheduler_placement,
                 include_source=include_source if include_source is not None else self._include_source_default,
-                _experimental_buffer_containers=_experimental_buffer_containers,
                 _experimental_proxy_ip=_experimental_proxy_ip,
                 _experimental_custom_scaling_factor=_experimental_custom_scaling_factor,
                 _experimental_enable_gpu_snapshot=_experimental_enable_gpu_snapshot,

modal/app.pyi

@@ -175,13 +175,14 @@ class _App:
         cpu: typing.Union[float, tuple[float, float], None] = None,
         memory: typing.Union[int, tuple[int, int], None] = None,
         ephemeral_disk: typing.Optional[int] = None,
+        min_containers: typing.Optional[int] = None,
+        max_containers: typing.Optional[int] = None,
+        buffer_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         proxy: typing.Optional[modal.proxy._Proxy] = None,
         retries: typing.Union[int, modal.retries.Retries, None] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
-        container_idle_timeout: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        keep_warm: typing.Optional[int] = None,
         name: typing.Optional[str] = None,
         is_generator: typing.Optional[bool] = None,
         cloud: typing.Optional[str] = None,
@@ -192,10 +193,13 @@ class _App:
         i6pn: typing.Optional[bool] = None,
         include_source: typing.Optional[bool] = None,
         _experimental_scheduler_placement: typing.Optional[modal.scheduler_placement.SchedulerPlacement] = None,
-        _experimental_buffer_containers: typing.Optional[int] = None,
         _experimental_proxy_ip: typing.Optional[str] = None,
         _experimental_custom_scaling_factor: typing.Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,
+        keep_warm: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
+        container_idle_timeout: typing.Optional[int] = None,
+        _experimental_buffer_containers: typing.Optional[int] = None,
     ) -> _FunctionDecoratorType: ...
     @typing_extensions.dataclass_transform(
         field_specifiers=(modal.cls.parameter,),
@@ -221,13 +225,14 @@ class _App:
         cpu: typing.Union[float, tuple[float, float], None] = None,
         memory: typing.Union[int, tuple[int, int], None] = None,
         ephemeral_disk: typing.Optional[int] = None,
+        min_containers: typing.Optional[int] = None,
+        max_containers: typing.Optional[int] = None,
+        buffer_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         proxy: typing.Optional[modal.proxy._Proxy] = None,
         retries: typing.Union[int, modal.retries.Retries, None] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
-        container_idle_timeout: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        keep_warm: typing.Optional[int] = None,
         cloud: typing.Optional[str] = None,
         region: typing.Union[str, collections.abc.Sequence[str], None] = None,
         enable_memory_snapshot: bool = False,
@@ -235,10 +240,13 @@ class _App:
         max_inputs: typing.Optional[int] = None,
         include_source: typing.Optional[bool] = None,
         _experimental_scheduler_placement: typing.Optional[modal.scheduler_placement.SchedulerPlacement] = None,
-        _experimental_buffer_containers: typing.Optional[int] = None,
         _experimental_proxy_ip: typing.Optional[str] = None,
         _experimental_custom_scaling_factor: typing.Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,
+        keep_warm: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
+        container_idle_timeout: typing.Optional[int] = None,
+        _experimental_buffer_containers: typing.Optional[int] = None,
     ) -> collections.abc.Callable[[CLS_T], CLS_T]: ...
     async def spawn_sandbox(
         self,
@@ -407,13 +415,14 @@ class App:
         cpu: typing.Union[float, tuple[float, float], None] = None,
         memory: typing.Union[int, tuple[int, int], None] = None,
         ephemeral_disk: typing.Optional[int] = None,
+        min_containers: typing.Optional[int] = None,
+        max_containers: typing.Optional[int] = None,
+        buffer_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         proxy: typing.Optional[modal.proxy.Proxy] = None,
         retries: typing.Union[int, modal.retries.Retries, None] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
-        container_idle_timeout: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        keep_warm: typing.Optional[int] = None,
         name: typing.Optional[str] = None,
         is_generator: typing.Optional[bool] = None,
         cloud: typing.Optional[str] = None,
@@ -424,10 +433,13 @@ class App:
         i6pn: typing.Optional[bool] = None,
         include_source: typing.Optional[bool] = None,
         _experimental_scheduler_placement: typing.Optional[modal.scheduler_placement.SchedulerPlacement] = None,
-        _experimental_buffer_containers: typing.Optional[int] = None,
         _experimental_proxy_ip: typing.Optional[str] = None,
         _experimental_custom_scaling_factor: typing.Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,
+        keep_warm: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
+        container_idle_timeout: typing.Optional[int] = None,
+        _experimental_buffer_containers: typing.Optional[int] = None,
     ) -> _FunctionDecoratorType: ...
     @typing_extensions.dataclass_transform(
         field_specifiers=(modal.cls.parameter,),
@@ -453,13 +465,14 @@ class App:
         cpu: typing.Union[float, tuple[float, float], None] = None,
         memory: typing.Union[int, tuple[int, int], None] = None,
         ephemeral_disk: typing.Optional[int] = None,
+        min_containers: typing.Optional[int] = None,
+        max_containers: typing.Optional[int] = None,
+        buffer_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         proxy: typing.Optional[modal.proxy.Proxy] = None,
         retries: typing.Union[int, modal.retries.Retries, None] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
-        container_idle_timeout: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        keep_warm: typing.Optional[int] = None,
         cloud: typing.Optional[str] = None,
         region: typing.Union[str, collections.abc.Sequence[str], None] = None,
         enable_memory_snapshot: bool = False,
@@ -467,10 +480,13 @@ class App:
         max_inputs: typing.Optional[int] = None,
         include_source: typing.Optional[bool] = None,
         _experimental_scheduler_placement: typing.Optional[modal.scheduler_placement.SchedulerPlacement] = None,
-        _experimental_buffer_containers: typing.Optional[int] = None,
         _experimental_proxy_ip: typing.Optional[str] = None,
         _experimental_custom_scaling_factor: typing.Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,
+        keep_warm: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
+        container_idle_timeout: typing.Optional[int] = None,
+        _experimental_buffer_containers: typing.Optional[int] = None,
     ) -> collections.abc.Callable[[CLS_T], CLS_T]: ...
 
     class __spawn_sandbox_spec(typing_extensions.Protocol[SUPERSELF]):

modal/cli/programs/vscode.py

@@ -82,7 +82,7 @@ def wait_for_port(data: tuple[str, str], q: Queue):
     timeout=args.get("timeout"),
     secrets=[Secret.from_dict({"MODAL_LAUNCH_ARGS": json.dumps(args)})],
     volumes=volumes,
-    concurrency_limit=1 if volume else None,
+    max_containers=1 if volume else None,
 )
 def run_vscode(q: Queue):
     os.chdir("/home/coder")

modal/client.pyi

@@ -27,7 +27,7 @@ class _Client:
     _snapshotted: bool
 
     def __init__(
-        self, server_url: str, client_type: int, credentials: typing.Optional[tuple[str, str]], version: str = "0.73.74"
+        self, server_url: str, client_type: int, credentials: typing.Optional[tuple[str, str]], version: str = "0.73.76"
     ): ...
     def is_closed(self) -> bool: ...
     @property
@@ -85,7 +85,7 @@ class Client:
     _snapshotted: bool
 
     def __init__(
-        self, server_url: str, client_type: int, credentials: typing.Optional[tuple[str, str]], version: str = "0.73.74"
+        self, server_url: str, client_type: int, credentials: typing.Optional[tuple[str, str]], version: str = "0.73.76"
     ): ...
     def is_closed(self) -> bool: ...
     @property

modal/cls.py

@@ -24,7 +24,7 @@ from ._resources import convert_fn_config_to_resources_config
 from ._serialization import check_valid_cls_constructor_arg
 from ._traceback import print_server_warnings
 from ._utils.async_utils import synchronize_api, synchronizer
-from ._utils.deprecation import deprecation_warning, renamed_parameter
+from ._utils.deprecation import deprecation_warning, renamed_parameter, warn_on_renamed_autoscaler_settings
 from ._utils.grpc_utils import retry_transient_errors
 from ._utils.mount_utils import validate_volumes
 from .client import _Client
@@ -559,6 +559,7 @@ class _Cls(_Object, type_prefix="cs"):
         cls._name = name
         return cls
 
+    @warn_on_renamed_autoscaler_settings
     def with_options(
         self: "_Cls",
         cpu: Optional[Union[float, tuple[float, float]]] = None,
@@ -567,10 +568,13 @@ class _Cls(_Object, type_prefix="cs"):
         secrets: Collection[_Secret] = (),
         volumes: dict[Union[str, os.PathLike], _Volume] = {},
         retries: Optional[Union[int, Retries]] = None,
+        max_containers: Optional[int] = None,  # Limit on the number of containers that can be concurrently running.
+        scaledown_window: Optional[int] = None,  # Max amount of time a container can remain idle before scaling down.
         timeout: Optional[int] = None,
-        concurrency_limit: Optional[int] = None,
         allow_concurrent_inputs: Optional[int] = None,
-        container_idle_timeout: Optional[int] = None,
+        # The following parameters are deprecated
+        concurrency_limit: Optional[int] = None,  # Now called `max_containers`
+        container_idle_timeout: Optional[int] = None,  # Now called `scaledown_window`
     ) -> "_Cls":
         """
         **Beta:** Allows for the runtime modification of a modal.Cls's configuration.
@@ -607,9 +611,10 @@ class _Cls(_Object, type_prefix="cs"):
             secret_ids=[secret.object_id for secret in secrets],
             resources=resources,
             retry_policy=retry_policy,
-            concurrency_limit=concurrency_limit,
+            # TODO(michael) Update the protos to use the new terminology
+            concurrency_limit=max_containers,
+            task_idle_timeout_secs=scaledown_window,
             timeout_secs=timeout,
-            task_idle_timeout_secs=container_idle_timeout,
             replace_volume_mounts=replace_volume_mounts,
             volume_mounts=volume_mounts,
             target_concurrent_inputs=allow_concurrent_inputs,

modal/cls.pyi

@@ -132,9 +132,11 @@ class _Cls(modal._object._Object):
         secrets: collections.abc.Collection[modal.secret._Secret] = (),
         volumes: dict[typing.Union[str, os.PathLike], modal.volume._Volume] = {},
         retries: typing.Union[int, modal.retries.Retries, None] = None,
+        max_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
         container_idle_timeout: typing.Optional[int] = None,
     ) -> _Cls: ...
     @staticmethod
@@ -191,9 +193,11 @@ class Cls(modal.object.Object):
         secrets: collections.abc.Collection[modal.secret.Secret] = (),
         volumes: dict[typing.Union[str, os.PathLike], modal.volume.Volume] = {},
         retries: typing.Union[int, modal.retries.Retries, None] = None,
+        max_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         timeout: typing.Optional[int] = None,
-        concurrency_limit: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
+        concurrency_limit: typing.Optional[int] = None,
         container_idle_timeout: typing.Optional[int] = None,
     ) -> Cls: ...
 

modal/config.py

@@ -60,6 +60,13 @@ Other possible configuration options are:
   When set, ignores the Image cache and builds all Image layers. Note that this
   will break the cache for all images based on the rebuilt layers, so other images
   may rebuild on subsequent runs / deploys even if the config is reverted.
+* `ignore_cache` (in the .toml file) / `MODAL_IGNORE_CACHE` (as an env var).
+  Defaults to False.
+  When set, ignores the Image cache and builds all Image layers. Unlike `force_build`,
+  this will not overwrite the cache for other images that have the same recipe.
+  Subsequent runs that do not use this option will pull the *previous* Image from
+  the cache, if one exists. It can be useful for testing an App's robustness to
+  Image rebuilds without clobbering Images used by other Apps.
 * `traceback` (in the .toml file) / `MODAL_TRACEBACK` (as an env var).
   Defaults to False. Enables printing full tracebacks on unexpected CLI
   errors, which can be useful for debugging client issues.
@@ -218,6 +225,7 @@ _SETTINGS = {
     "worker_id": _Setting(),  # For internal debugging use.
     "restore_state_path": _Setting("/__modal/restore-state.json"),
     "force_build": _Setting(False, transform=_to_boolean),
+    "ignore_cache": _Setting(False, transform=_to_boolean),
     "traceback": _Setting(False, transform=_to_boolean),
     "image_builder_version": _Setting(),
     "strict_parameters": _Setting(False, transform=_to_boolean),  # For internal/experimental use

modal/functions.pyi

@@ -73,17 +73,18 @@ class Function(
             typing.Union[modal.volume.Volume, modal.cloud_bucket_mount.CloudBucketMount],
         ] = {},
         webhook_config: typing.Optional[modal_proto.api_pb2.WebhookConfig] = None,
+        cpu: typing.Union[float, tuple[float, float], None] = None,
         memory: typing.Union[int, tuple[int, int], None] = None,
         proxy: typing.Optional[modal.proxy.Proxy] = None,
         retries: typing.Union[int, modal.retries.Retries, None] = None,
         timeout: typing.Optional[int] = None,
-        concurrency_limit: typing.Optional[int] = None,
+        min_containers: typing.Optional[int] = None,
+        max_containers: typing.Optional[int] = None,
+        buffer_containers: typing.Optional[int] = None,
+        scaledown_window: typing.Optional[int] = None,
         allow_concurrent_inputs: typing.Optional[int] = None,
         batch_max_size: typing.Optional[int] = None,
         batch_wait_ms: typing.Optional[int] = None,
-        container_idle_timeout: typing.Optional[int] = None,
-        cpu: typing.Union[float, tuple[float, float], None] = None,
-        keep_warm: typing.Optional[int] = None,
         cloud: typing.Optional[str] = None,
         scheduler_placement: typing.Optional[modal.scheduler_placement.SchedulerPlacement] = None,
         is_builder_function: bool = False,
@@ -95,7 +96,6 @@ class Function(
         max_inputs: typing.Optional[int] = None,
         ephemeral_disk: typing.Optional[int] = None,
         include_source: typing.Optional[bool] = None,
-        _experimental_buffer_containers: typing.Optional[int] = None,
         _experimental_proxy_ip: typing.Optional[str] = None,
         _experimental_custom_scaling_factor: typing.Optional[float] = None,
         _experimental_enable_gpu_snapshot: bool = False,

modal/image.py

@@ -622,6 +622,7 @@ class _Image(_Object, type_prefix="im"):
                 # Failsafe mechanism to prevent inadvertant updates to the global images.
                 # Only admins can publish to the global namespace, but they have to additionally request it.
                 allow_global_deployment=os.environ.get("MODAL_IMAGE_ALLOW_GLOBAL_DEPLOYMENT", "0") == "1",
+                ignore_cache=config.get("ignore_cache"),
             )
             resp = await retry_transient_errors(resolver.client.stub.ImageGetOrCreate, req)
             image_id = resp.image_id

{modal-0.73.74.dist-info → modal-0.73.76.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: modal
-Version: 0.73.74
+Version: 0.73.76
 Summary: Python client library for Modal
 Author-email: Modal Labs <support@modal.com>
 License: Apache-2.0

{modal-0.73.74.dist-info → modal-0.73.76.dist-info}/RECORD

@@ -3,7 +3,7 @@ modal/__main__.py,sha256=CgIjP8m1xJjjd4AXc-delmR6LdBCZclw2A_V38CFIio,2870
 modal/_clustered_functions.py,sha256=kTf-9YBXY88NutC1akI-gCbvf01RhMPCw-zoOI_YIUE,2700
 modal/_clustered_functions.pyi,sha256=vllkegc99A0jrUOWa8mdlSbdp6uz36TsHhGxysAOpaQ,771
 modal/_container_entrypoint.py,sha256=arhkIoF8nQNfa4iwYGSoqN3QMDg5M38QNAODXC8TlKc,29301
-modal/_functions.py,sha256=34kHxFM4k1oSb0OV2Op6uDuiT-jEBG8cbo3NfqhAgew,71417
+modal/_functions.py,sha256=LkzRfcex420bHWN2s5Aqbtm9cg7PaAcuG-H2jNz1JfE,71459
 modal/_ipython.py,sha256=TW1fkVOmZL3YYqdS2YlM1hqpf654Yf8ZyybHdBnlhSw,301
 modal/_location.py,sha256=joiX-0ZeutEUDTrrqLF1GHXCdVLF-rHzstocbMcd_-k,366
 modal/_object.py,sha256=ItQcsMNkz9Y3kdTsvfNarbW-paJ2qabDyQ7njaqY0XI,11359
@@ -18,16 +18,16 @@ modal/_traceback.py,sha256=IZQzB3fVlUfMHOSyKUgw0H6qv4yHnpyq-XVCNZKfUdA,5023
 modal/_tunnel.py,sha256=zTBxBiuH1O22tS1OliAJdIsSmaZS8PlnifS_6S5z-mk,6320
 modal/_tunnel.pyi,sha256=JmmDYAy9F1FpgJ_hWx0xkom2nTOFQjn4mTPYlU3PFo4,1245
 modal/_watcher.py,sha256=K6LYnlmSGQB4tWWI9JADv-tvSvQ1j522FwT71B51CX8,3584
-modal/app.py,sha256=o5mHoHtn41nkvskX_ekJkyfG6MXwj5rqerRi_nnPd0w,44725
-modal/app.pyi,sha256=0MMCgskIL4r3eq8oBcfm2lLyeao2gXjS3iXaIfmaJ-o,25959
+modal/app.py,sha256=kF3frIt4eRKVYYCjusMMhKJpO_lDdm2z37HOXPwpjT8,45506
+modal/app.pyi,sha256=tZFbcsu20SuvfB2puxCyuXLFNJ9bQulzag55rVpgZmc,26827
 modal/call_graph.py,sha256=1g2DGcMIJvRy-xKicuf63IVE98gJSnQsr8R_NVMptNc,2581
 modal/client.py,sha256=8SQawr7P1PNUCq1UmJMUQXG2jIo4Nmdcs311XqrNLRE,15276
-modal/client.pyi,sha256=PwDnBFBHB369Q6eY1uz7_LblIn1IN9DqaTm0pAhpEGU,7593
+modal/client.pyi,sha256=BY_85JYs5yMmzZbJ4XMfajA48crezjBUTGtZ8RUi-gc,7593
 modal/cloud_bucket_mount.py,sha256=YOe9nnvSr4ZbeCn587d7_VhE9IioZYRvF9VYQTQux08,5914
 modal/cloud_bucket_mount.pyi,sha256=30T3K1a89l6wzmEJ_J9iWv9SknoGqaZDx59Xs-ZQcmk,1607
-modal/cls.py,sha256=nx3mFniMw_b0tKqmY5smbi39xSpwmu_Mnb6_h1OFIBM,30208
-modal/cls.pyi,sha256=4Ms1i4Wty1qe49Dh_wsGhJDCiJz7t-XGqXLcpzwhUqs,9084
-modal/config.py,sha256=E1bCdwNziBorqXal6Cble7gfLNzGqSuEgprRJwzYqg4,11160
+modal/cls.py,sha256=pQqLU_vlgHD2aglDQAoVuzt0BS3pUkRfC7oHG_fsbV4,30700
+modal/cls.pyi,sha256=MAx5J72UhD5l2qbH_sTpZDP34PlI4P71OBLSWOqvE68,9300
+modal/config.py,sha256=Boz1bPzaG-k5Grjq6y6fAELH1N_gTuYDnpB6FODzCPo,11710
 modal/container_process.py,sha256=WTqLn01dJPVkPpwR_0w_JH96ceN5mV4TGtiu1ZR2RRA,6108
 modal/container_process.pyi,sha256=Hf0J5JyDdCCXBJSKx6gvkPOo0XrztCm78xzxamtzUjQ,2828
 modal/dict.py,sha256=vc5lQVqzeDUCb4fRjnOlqYK2GmBb0fIhZmvB0xIBG0U,12921
@@ -41,9 +41,9 @@ modal/file_io.py,sha256=lcMs_E9Xfm0YX1t9U2wNIBPnqHRxmImqjLW1GHqVmyg,20945
 modal/file_io.pyi,sha256=NTRft1tbPSWf9TlWVeZmTlgB5AZ_Zhu2srWIrWr7brk,9445
 modal/file_pattern_matcher.py,sha256=trosX-Bp7dOubudN1bLLhRAoidWy1TcoaR4Pv8CedWw,6497
 modal/functions.py,sha256=kcNHvqeGBxPI7Cgd57NIBBghkfbeFJzXO44WW0jSmao,325
-modal/functions.pyi,sha256=2z5Vt2U3KfmTWyC55hPVEgc4i83BiqAK0faJgumpe04,14405
+modal/functions.pyi,sha256=7m3qSR7dKUPcSZycZ4Y1VaySjdBZI0_MmR8pRrbHzxE,14387
 modal/gpu.py,sha256=Kbhs_u49FaC2Zi0TjCdrpstpRtT5eZgecynmQi5IZVE,6752
-modal/image.py,sha256=Bs1ND2WkLr9CBtj0heO7e3w9uGCnijKU8owiQSRQXv0,90200
+modal/image.py,sha256=adMUpS7WrCu-M78BWslz2r6GPviy4qPvd5Dh-dBIrrk,90257
 modal/image.pyi,sha256=L7aZUOElSGtNHmFHz1RgKP1cG5paiXt_EzylrwBwzVk,25004
 modal/io_streams.py,sha256=QkQiizKRzd5bnbKQsap31LJgBYlAnj4-XkV_50xPYX0,15079
 modal/io_streams.pyi,sha256=bJ7ZLmSmJ0nKoa6r4FJpbqvzdUVa0lEe0Fa-MMpMezU,5071
@@ -96,7 +96,7 @@ modal/_utils/app_utils.py,sha256=88BT4TPLWfYAQwKTHcyzNQRHg8n9B-QE2UyJs96iV-0,108
 modal/_utils/async_utils.py,sha256=5PdDuI1aSwPOI4a3dIvW0DkPqGw6KZN6RtWE18Dzv1E,25079
 modal/_utils/blob_utils.py,sha256=RB1G6T7eC1Poe-O45qYLaxwCr2jkM-Q6Nexk1J3wk_w,14505
 modal/_utils/bytes_io_segment_payload.py,sha256=uunxVJS4PE1LojF_UpURMzVK9GuvmYWRqQo_bxEj5TU,3385
-modal/_utils/deprecation.py,sha256=dycySRBxyZf3ITzEqPNM6MxXTk9-0VVLA8oCPQ5j_Os,3426
+modal/_utils/deprecation.py,sha256=rgCGTrk-u_uaDXNDTAW9FM8GP8N3ErlDfr2wXhKYLVw,4870
 modal/_utils/docker_utils.py,sha256=h1uETghR40mp_y3fSWuZAfbIASH1HMzuphJHghAL6DU,3722
 modal/_utils/function_utils.py,sha256=Rmz8GJDie-RW_q2RcTwholEWixS2IQDPBsRBJ3f3ZvU,27302
 modal/_utils/grpc_testing.py,sha256=H1zHqthv19eGPJz2HKXDyWXWGSqO4BRsxah3L5Xaa8A,8619
@@ -135,7 +135,7 @@ modal/cli/utils.py,sha256=hZmjyzcPjDnQSkLvycZD2LhGdcsfdZshs_rOU78EpvI,3717
 modal/cli/volume.py,sha256=c2IuVNO2yJVaXmZkRh3xwQmznlRTgFoJr_BIzzqtVv0,10251
 modal/cli/programs/__init__.py,sha256=svYKtV8HDwDCN86zbdWqyq5T8sMdGDj0PVlzc2tIxDM,28
 modal/cli/programs/run_jupyter.py,sha256=MX6YQ6zRyRk1xo8tYZFiGam0p5KETwax81L6TpaS9I0,2778
-modal/cli/programs/vscode.py,sha256=Q0FA0PBrGf8SVlokenFhcGg-Viaj1oKSuhHV7m35CwQ,3476
+modal/cli/programs/vscode.py,sha256=kfvhZQ4bJwtVm3MgC1V7AlygZOlKT1a33alr_uwrewA,3473
 modal/extensions/__init__.py,sha256=waLjl5c6IPDhSsdWAm9Bji4e2PVxamYABKAze6CHVXY,28
 modal/extensions/ipython.py,sha256=Xvzy-A7cvwMSDa9p4c4CEMLOX2_Xsg9DkM1J9uyu7jc,983
 modal/requirements/2023.12.312.txt,sha256=zWWUVgVQ92GXBKNYYr2-5vn9rlnXcmkqlwlX5u1eTYw,400
@@ -168,10 +168,10 @@ modal_proto/options_pb2_grpc.pyi,sha256=CImmhxHsYnF09iENPoe8S4J-n93jtgUYD2JPAc0y
 modal_proto/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 modal_version/__init__.py,sha256=wiJQ53c-OMs0Xf1UeXOxQ7FwlV1VzIjnX6o-pRYZ_Pk,470
 modal_version/__main__.py,sha256=2FO0yYQQwDTh6udt1h-cBnGd1c4ZyHnHSI4BksxzVac,105
-modal_version/_version_generated.py,sha256=EHhRz20OtH8bXhxnn9hwOhBRPymrrkxTY3fwiBGngnk,149
-modal-0.73.74.dist-info/LICENSE,sha256=psuoW8kuDP96RQsdhzwOqi6fyWv0ct8CR6Jr7He_P_k,10173
-modal-0.73.74.dist-info/METADATA,sha256=Wvg_qdimqEI33t3EKm1VaW419bcX8Fwkith8J7nnCnA,2452
-modal-0.73.74.dist-info/WHEEL,sha256=jB7zZ3N9hIM9adW7qlTAyycLYW9npaWKLRzaoVcLKcM,91
-modal-0.73.74.dist-info/entry_points.txt,sha256=An-wYgeEUnm6xzrAP9_NTSTSciYvvEWsMZILtYrvpAI,46
-modal-0.73.74.dist-info/top_level.txt,sha256=4BWzoKYREKUZ5iyPzZpjqx4G8uB5TWxXPDwibLcVa7k,43
-modal-0.73.74.dist-info/RECORD,,
+modal_version/_version_generated.py,sha256=jmVn-W1r2Lr0q91iDTY2ZmvlYDwd2coi9KKtw_i03Jk,149
+modal-0.73.76.dist-info/LICENSE,sha256=psuoW8kuDP96RQsdhzwOqi6fyWv0ct8CR6Jr7He_P_k,10173
+modal-0.73.76.dist-info/METADATA,sha256=XZV1Q5CvoLdg2NmRCtLXu5CfAd8SISK_6cRFO_bySec,2452
+modal-0.73.76.dist-info/WHEEL,sha256=jB7zZ3N9hIM9adW7qlTAyycLYW9npaWKLRzaoVcLKcM,91
+modal-0.73.76.dist-info/entry_points.txt,sha256=An-wYgeEUnm6xzrAP9_NTSTSciYvvEWsMZILtYrvpAI,46
+modal-0.73.76.dist-info/top_level.txt,sha256=4BWzoKYREKUZ5iyPzZpjqx4G8uB5TWxXPDwibLcVa7k,43
+modal-0.73.76.dist-info/RECORD,,

modal_version/_version_generated.py

@@ -1,4 +1,4 @@
 # Copyright Modal Labs 2025
 
 # Note: Reset this value to -1 whenever you make a minor `0.X` release of the client.
-build_number = 74  # git: bdcac62
+build_number = 76  # git: d55968d