[jira] [Comment Edited] (FLINK-19687) Support to get execution plan in `StatementSet`

["xiaozilong (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/FLINK-19687?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17215827#comment-17215827 ] 

xiaozilong edited comment on FLINK-19687 at 10/17/20, 9:37 AM:
---------------------------------------------------------------

Hi [~hailong wang], I want to visualize the execution plan in [https://flink.apache.org/visualizer/] tools. The physical execution plan unable to draw. In addition, I think it is allowed to get the execution plan before the job is submitted in Flink SQL 1.11.


was (Author: xiaozilong):
Hi [~hailong wang], I want to visualize the execution plan in [https://flink.apache.org/visualizer/] tools. The physical execution plan unable to draw. In addition, I think it is allowed to get the execution plan before the job is submitted.

> Support to get execution plan in `StatementSet`
> -----------------------------------------------
>
>                 Key: FLINK-19687
>                 URL: https://issues.apache.org/jira/browse/FLINK-19687
>             Project: Flink
>          Issue Type: Wish
>          Components: Table SQL / API
>    Affects Versions: 1.11.0
>            Reporter: xiaozilong
>            Priority: Major
>
> Hi, I want to get job's execution plan in Flink SQL 1.11, but i meet exception "No operators defined in streaming topology. Cannot execute." when use `env.getExecutionPlan()`. The same code runs fine in Flink SQL 1.10. I found translation operations only happen when StatementSet.execute() is called in Flink SQL 1.11. So we cannot get job's execution plan before the job submit? Can we support to get execution plan in `StatementSet` or invoke method `TableEnvironmentImpl#translate` in `StatementSetImpl#addInsertSql`? I think the latter is better so that we can reuse `env.getExecutionPlan()`.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)