Retry pulling from the queue when DatabaseOperationalError is raised

The Queue.pull method can raise such exception even with an autocommit
transaction because the update query can be in conflict with another same
update query. This is because we use repeatable isolation level.

Closes #11937

trytond/trytond/worker.py

@@ -75,7 +75,10 @@ def work(options):
             while len(tasks.filter()) >= processes:
                 time.sleep(0.1)
             for queue in queues:
-                task_id, next_ = queue.pull(options.name)
+                try:
+                    task_id, next_ = queue.pull(options.name)
+                except backend.DatabaseOperationalError:
+                    break
                 if next_ is not None:
                     timeout = min(next_, timeout)
                 if task_id: