[
https://issues.apache.org/jira/browse/SPARK-33038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17204958#comment-17204958
]
Apache Spark commented on SPARK-33038:
--------------------------------------
User 'allisonwang-db' has created a pull request for this issue:
https://github.com/apache/spark/pull/29915
> AQE plan string should only display one plan when the initial and the current
> plan are the same
> -----------------------------------------------------------------------------------------------
>
> Key: SPARK-33038
> URL: https://issues.apache.org/jira/browse/SPARK-33038
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.1.0
> Reporter: Allison Wang
> Assignee: Apache Spark
> Priority: Minor
>
> Currently, the AQE plan string displays both the initial plan and the current
> or the final plan. This can be redundant when the initial plan and the
> current physical plan are exactly the same. For instance, the `EXPLAIN`
> command will not actually execute the query, and thus the plan string will
> never change, but currently, the plan string still shows both the current and
> the initial plan:
>
> {code:java}
> AdaptiveSparkPlan (8)
> +- == Current Plan ==
> Sort (7)
> +- Exchange (6)
> +- HashAggregate (5)
> +- Exchange (4)
> +- HashAggregate (3)
> +- Filter (2)
> +- Scan parquet default.explain_temp1 (1)
> +- == Initial Plan ==
> Sort (7)
> +- Exchange (6)
> +- HashAggregate (5)
> +- Exchange (4)
> +- HashAggregate (3)
> +- Filter (2)
> +- Scan parquet default.explain_temp1 (1)
> {code}
> When the initial and the current plan are the same, there should be only one
> plan string displayed. For example
> {code:java}
> AdaptiveSparkPlan (8)
> +- Sort (7)
> +- Exchange (6)
> +- HashAggregate (5)
> +- Exchange (4)
> +- HashAggregate (3)
> +- Filter (2)
> +- Scan parquet default.explain_temp1 (1){code}
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
