Log when evaluator is leaving events

The batching interval and size can maybe be tuned slightly, before that
it will be quite interesting to know how often message processing
is postponed.

This logging feature existed in the sync version of the ensemble
evaluator, but was not carried over in 509c224.

src/ert/ensemble_evaluator/evaluator.py

@@ -149,7 +149,14 @@ def set_event_handler(event_types: Set[Type[Event]], func: Any) -> None:
                 except asyncio.TimeoutError:
                     continue
             self._complete_batch.set()
+            batch_start_time = asyncio.get_running_loop().time()
             await self._batch_processing_queue.put(batch)
+            if batch and self._events.qsize() > 0:
+                logger.info(
+                    f"Processed {len(batch)} events in "
+                    f"{(asyncio.get_running_loop().time() - batch_start_time):.6f}s "
+                    f"{self._events.qsize()} left in queue"
+                )
 
     async def _fm_handler(
         self, events: Sequence[Union[FMEvent, RealizationEvent]]