[jira] [Reopened] (FLINK-2249) ExecutionEnvironment: Ignore calls to execute() if no data sinks defined

["Maximilian Michels (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/FLINK-2249?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Maximilian Michels reopened FLINK-2249:
---------------------------------------

> ExecutionEnvironment: Ignore calls to execute() if no data sinks defined
> ------------------------------------------------------------------------
>
>                 Key: FLINK-2249
>                 URL: https://issues.apache.org/jira/browse/FLINK-2249
>             Project: Flink
>          Issue Type: Improvement
>          Components: Java API, Scala API
>    Affects Versions: 0.9
>            Reporter: Maximilian Michels
>            Assignee: Chesnay Schepler
>
> The basic skeleton of a Flink program looks like this: 
> {code}
> ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
> // bootstrap DataSet
> DataSet<..> ds = env.fromElements(1,2,3,4);
> // perform transformations
> ..
> // define sinks, e.g.
> ds.writeToTextFile("/some/path");
> // execute
> env.execute()
> {code}
> First thing users do is to change {{ds.writeToTextFile("/some/path");}} into {{ds.print();}}. But that fails with an Exception ("No new data sinks defined...").
> In FLINK-2026 we made this exception message easier to understand. However, users still don't understand what is happening. Especially because they see Flink executing and then failing.
> I propose to ignore calls to execute() when no sinks are defined. Instead, we should just print a warning: "Detected call to execute without any data sinks. Not executing."



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)