[SPARK-30444] The same job will be computated for many times when using Dataset.show()

• Type: Improvement
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.1.0
• Component/s: SQL
• Assignee: Unassigned
• Reporter: IcySanwitch
• Created: 07/Jan/20 01:55
• Updated: 12/Dec/22 18:11

Description

When I run the example sql.SparkSQLExample, df.show() at line 60 would trigger an action. On WebUI, I noticed that this API creates 5 jobs, all of which have the same lineage graph with the same RDDs and the same call stacks. That means Spark recomputates the job for 5 times. But strangely, sqlDF.show() at line 123 only creates 1 job.

I don't know what happened at show() at line 60.