[jira] [Closed] (FLINK-2067) Chained streaming operators should not throw chained exceptions

["Stephan Ewen (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/FLINK-2067?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stephan Ewen closed FLINK-2067.
-------------------------------

> Chained streaming operators should not throw chained exceptions
> ---------------------------------------------------------------
>
>                 Key: FLINK-2067
>                 URL: https://issues.apache.org/jira/browse/FLINK-2067
>             Project: Flink
>          Issue Type: Bug
>          Components: DataStream API
>            Reporter: Stephan Ewen
>            Assignee: Stephan Ewen
>             Fix For: 1.3.0
>
>
> The exceptions that come from chained operators have an non-intuitive chaining structure, that makes the stack traces harder to understand.
> For every chained task, there is a "Failed to forward record" exception, before the actual exception comes as a cause.
> In the Batch API, we use a special "ExceptionInChainedStubException" that is recognized and un-nested to make chained operator exceptions surface as root exceptions. 
> We should do the same for the streaming API.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)