sentry-arroyo 2.29.6__tar.gz → 2.29.7__tar.gz

{sentry_arroyo-2.29.6/sentry_arroyo.egg-info → sentry_arroyo-2.29.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sentry-arroyo
-Version: 2.29.6
+Version: 2.29.7
 Summary: Arroyo is a Python library for working with streaming data.
 Home-page: https://github.com/getsentry/arroyo
 Author: Sentry

{sentry_arroyo-2.29.6 → sentry_arroyo-2.29.7}/arroyo/processing/strategies/run_task_with_multiprocessing.py

@@ -41,6 +41,7 @@ DEFAULT_INPUT_BLOCK_SIZE = 16 * 1024 * 1024
 DEFAULT_OUTPUT_BLOCK_SIZE = 16 * 1024 * 1024
 
 LOG_THRESHOLD_TIME = 20  # In seconds
+FAIL_THRESHOLD_TIME = 180  # In seconds
 
 
 class ChildProcessTerminated(RuntimeError):
@@ -600,6 +601,7 @@ class RunTaskWithMultiprocessing(
             "arroyo.strategies.run_task_with_multiprocessing.batches_in_progress",
         )
         self.__pool_waiting_time: Optional[float] = None
+        self.__pool_waiting_log_time: Optional[float] = None
         self.__metrics.gauge(
             "arroyo.strategies.run_task_with_multiprocessing.processes", num_processes
         )
@@ -673,22 +675,31 @@ class RunTaskWithMultiprocessing(
                 if deadline is None or deadline > time.time():
                     continue
                 break
-            except multiprocessing.TimeoutError:
-                if self.__pool_waiting_time is None:
-                    self.__pool_waiting_time = time.time()
+            except multiprocessing.TimeoutError as err:
+                current_time = time.time()
+                if (
+                    self.__pool_waiting_time is None
+                    or self.__pool_waiting_log_time is None
+                ):
+                    self.__pool_waiting_time = current_time
+                    self.__pool_waiting_log_time = current_time
                 else:
-                    current_time = time.time()
-                    if current_time - self.__pool_waiting_time > LOG_THRESHOLD_TIME:
+                    if current_time - self.__pool_waiting_log_time > LOG_THRESHOLD_TIME:
                         logger.warning(
                             "Waited on the process pool longer than %d seconds. Waiting for %d results. Pool: %r",
                             LOG_THRESHOLD_TIME,
                             len(self.__processes),
                             self.__pool,
                         )
-                        self.__pool_waiting_time = current_time
+                        self.__pool_waiting_log_time = current_time
+                    if current_time - self.__pool_waiting_time > FAIL_THRESHOLD_TIME:
+                        raise multiprocessing.TimeoutError(
+                            f"Waited on process pool for {FAIL_THRESHOLD_TIME} seconds giving up."
+                        ) from err
                 break
             else:
                 self.__pool_waiting_time = None
+                self.__pool_waiting_log_time = None
 
     def __check_for_results_impl(self, timeout: Optional[float] = None) -> None:
         (

{sentry_arroyo-2.29.6 → sentry_arroyo-2.29.7/sentry_arroyo.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sentry-arroyo
-Version: 2.29.6
+Version: 2.29.7
 Summary: Arroyo is a Python library for working with streaming data.
 Home-page: https://github.com/getsentry/arroyo
 Author: Sentry

{sentry_arroyo-2.29.6 → sentry_arroyo-2.29.7}/setup.py

@@ -10,7 +10,7 @@ def get_requirements() -> Sequence[str]:
 
 setup(
     name="sentry-arroyo",
-    version="2.29.6",
+    version="2.29.7",
     author="Sentry",
     author_email="oss@sentry.io",
     license="Apache-2.0",