[GitHub] [airflow] ericpollmann edited a comment on issue #21082: Scheduler hangs, DAG serialization error under high scheduler load

[GitBox <gi...@apache.org>]

ericpollmann edited a comment on issue #21082:
URL: https://github.com/apache/airflow/issues/21082#issuecomment-1020795260


   This makes the issue go away but seems to be addressing the symptom rather than the cause - not sure how an unprocessed python dict is reaching this code that assumes params has been processed into a specific format.
   
   ```
   diff --git a/airflow/serialization/serialized_objects.py b/airflow/serialization/serialized_objects.py
   index 63820ffdf..f165b454b 100644
   --- a/airflow/serialization/serialized_objects.py
   +++ b/airflow/serialization/serialized_objects.py
   @@ -466,7 +466,9 @@ class BaseSerialization:
            """Serialize Params dict for a DAG/Task"""
            serialized_params = {}
            for k, v in params.items():
   -            # TODO: As of now, we would allow serialization of params which are of type Param only.
   +            # Old style params, convert it
   +            if not isinstance(v, Param):
   +                v = Param(v)
                try:
                    class_identity = f"{v.__module__}.{v.__class__.__name__}"
                except AttributeError:
   ```
   
   This may potentially be happening in some place like this?
   https://github.com/apache/airflow/blob/602abe8394fafe7de54df7e73af56de848cdf617/airflow/models/taskinstance.py#L1988


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@airflow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org