Details
-
Improvement
-
Status: Open
-
P3
-
Resolution: Unresolved
-
None
-
None
Description
Currently, job servers are shut down when the Python script exits [1]. A better long-term solution would be to shut them down instead when the pipeline is finished executing, such as [2]. This will put resource management in a common code path that is less error-prone.
Attachments
Issue Links
- is related to
-
BEAM-7973 Python doesn't shut down Flink job server properly
- Resolved