Details
-
Wish
-
Status: Open
-
Major
-
Resolution: Unresolved
-
2.1.3
-
None
-
None
Description
Backgroud:
In industry development environment, we got used to write several spark-sql blocks in one pyspark script. Without really submitting to cluster as running application, getting all the physical plans of a application is too hard to do within a parameter statement of "spark-submit" command.
Wish:
I wish to add a parameter "–genPlan" as for "spark-submit" command, which can return all the physical plans of a application instead of submitting to run indeed. Also, other approach to settle the matter is wellcome.