fix: Set a timeout for Actor cleanup (#206)

- closes #200

I could not find the actual reason why the linked run got stuck, but
this should make the Actor cleanup more robust as a whole.

CHANGELOG.md

@@ -2,7 +2,9 @@
 
 ## [1.7.1](../../releases/tag/v1.7.1) - Unreleased
 
-...
+### Fixed
+
+- Set a timeout for Actor cleanup
 
 ## [1.7.0](../../releases/tag/v1.7.0) - 2024-03-12
 

src/apify/actor.py

@@ -5,7 +5,7 @@
 import inspect
 import os
 import sys
-from datetime import datetime, timezone
+from datetime import datetime, timedelta, timezone
 from typing import TYPE_CHECKING, Any, Awaitable, Callable, TypeVar, cast
 
 from apify_client import ApifyClientAsync
@@ -301,6 +301,7 @@ async def exit(
         exit_code: int = 0,
         event_listeners_timeout_secs: float | None = EVENT_LISTENERS_TIMEOUT_SECS,
         status_message: str | None = None,
+        cleanup_timeout: timedelta = timedelta(seconds=30),
     ) -> None:
         """Exit the actor instance.
 
@@ -314,11 +315,13 @@ async def exit(
             exit_code (int, optional): The exit code with which the actor should fail (defaults to `0`).
             event_listeners_timeout_secs (float, optional): How long should the actor wait for actor event listeners to finish before exiting.
             status_message (str, optional): The final status message that the actor should display.
+            cleanup_timeout (timedelta, optional): How long we should wait for event listeners.
         """
         return await cls._get_default_instance().exit(
             exit_code=exit_code,
             event_listeners_timeout_secs=event_listeners_timeout_secs,
             status_message=status_message,
+            cleanup_timeout=cleanup_timeout,
         )
 
     async def _exit_internal(
@@ -327,6 +330,7 @@ async def _exit_internal(
         exit_code: int = 0,
         event_listeners_timeout_secs: float | None = EVENT_LISTENERS_TIMEOUT_SECS,
         status_message: str | None = None,
+        cleanup_timeout: timedelta = timedelta(seconds=30),
     ) -> None:
         self._raise_if_not_initialized()
 
@@ -336,21 +340,23 @@ async def _exit_internal(
 
         self.log.info('Exiting actor', extra={'exit_code': exit_code})
 
-        await self._cancel_event_emitting_intervals()
+        async def finalize() -> None:
+            await self._cancel_event_emitting_intervals()
 
-        # Send final persist state event
-        if not self._was_final_persist_state_emitted:
-            self._event_manager.emit(ActorEventTypes.PERSIST_STATE, {'isMigrating': False})
-            self._was_final_persist_state_emitted = True
+            # Send final persist state event
+            if not self._was_final_persist_state_emitted:
+                self._event_manager.emit(ActorEventTypes.PERSIST_STATE, {'isMigrating': False})
+                self._was_final_persist_state_emitted = True
 
-        if status_message is not None:
-            await self.set_status_message(status_message, is_terminal=True)
+            if status_message is not None:
+                await self.set_status_message(status_message, is_terminal=True)
 
-        # Sleep for a bit so that the listeners have a chance to trigger
-        await asyncio.sleep(0.1)
+            # Sleep for a bit so that the listeners have a chance to trigger
+            await asyncio.sleep(0.1)
 
-        await self._event_manager.close(event_listeners_timeout_secs=event_listeners_timeout_secs)
+            await self._event_manager.close(event_listeners_timeout_secs=event_listeners_timeout_secs)
 
+        await asyncio.wait_for(finalize(), cleanup_timeout.total_seconds())
         self._is_initialized = False
 
         if is_running_in_ipython():