Fix: Handle dead consumer threads in Google Cloud Functions environment

langfuse/_client/resource_manager.py

@@ -408,7 +408,53 @@ def flush(self) -> None:
         self._score_ingestion_queue.join()
         langfuse_logger.debug("Successfully flushed score ingestion queue")
 
-        self._media_upload_queue.join()
+        # Check if threads are alive AND healthy (recently active)
+        healthy_threads = [
+            c for c in self._media_upload_consumers
+            if c.is_alive() and c.is_healthy(timeout_seconds=5.0)
+        ]
+
+        if healthy_threads:
+            # Wait for queue to be processed, but with a timeout
+            langfuse_logger.debug(
+                f"{len(healthy_threads)} healthy consumer threads active, waiting for queue to drain"
+            )
+            start_time = time.time()
+            timeout = 30
+
+            while not self._media_upload_queue.empty() and (time.time() - start_time) < timeout:
+                time.sleep(0.1)
+
+            if self._media_upload_queue.empty():
+                langfuse_logger.debug("Successfully flushed media upload queue via consumer threads")
+                return
+            else:
+                langfuse_logger.warning(
+                    f"Media upload queue not empty after {timeout}s, "
+                    f"{self._media_upload_queue.qsize()} items remaining. "
+                    f"Processing synchronously."
+                )
+        else:
+            alive_count = len([c for c in self._media_upload_consumers if c.is_alive()])
+            langfuse_logger.warning(
+                f"Consumer threads unhealthy or dead ({alive_count} alive but unhealthy). "
+                f"Processing {self._media_upload_queue.qsize()} queued items synchronously."
+            )
+
+        # Synchronous fallback processing
+        items_processed = 0
+        while not self._media_upload_queue.empty():
+            try:
+                self._media_manager.process_next_media_upload()
+                items_processed += 1
+            except Exception as e:
+                langfuse_logger.error(f"Error processing media upload synchronously: {e}")
+
+        if items_processed > 0:
+            langfuse_logger.info(
+                f"Processed {items_processed} media uploads synchronously in flush()"
+            )
+
         langfuse_logger.debug("Successfully flushed media upload queue")
 
     def shutdown(self) -> None:

langfuse/_task_manager/media_upload_consumer.py

@@ -1,5 +1,6 @@
 import logging
 import threading
+import time
 
 from .media_manager import MediaManager
 
@@ -25,6 +26,8 @@ def __init__(
         # run() *after* we set it to False in pause... and keep running
         # forever.
         self.running = True
+        # Track when thread last processed something
+        self.last_activity = time.time()
         self._identifier = identifier
         self._media_manager = media_manager
 
@@ -34,11 +37,33 @@ def run(self) -> None:
             f"Thread: Media upload consumer thread #{self._identifier} started and actively processing queue items"
         )
         while self.running:
-            self._media_manager.process_next_media_upload()
+            try:
+                # Update activity timestamp before processing
+                self.last_activity = time.time()
+                self._media_manager.process_next_media_upload()
+                # Update after successful processing
+                self.last_activity = time.time()
+            except Exception as e:
+                self._log.error(
+                    f"Thread #{self._identifier}: Unexpected error in consumer loop: {e}"
+                )
+                # Continue running despite errors
+                time.sleep(0.1)
 
     def pause(self) -> None:
         """Pause the media upload consumer."""
         self._log.debug(
             f"Thread: Pausing media upload consumer thread #{self._identifier}"
         )
         self.running = False
+
+    def is_healthy(self, timeout_seconds: float = 5.0) -> bool:
+        """
+        Check if thread is healthy and recently active.
+        Returns False if thread hasn't processed anything in timeout_seconds.
+        """
+        if not self.is_alive():
+            return False
+
+        time_since_activity = time.time() - self.last_activity
+        return time_since_activity < timeout_seconds