[jira] [Commented] (NIFI-9825) Execution "Primary node" with incoming connections limitation bug

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

    [ https://issues.apache.org/jira/browse/NIFI-9825?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17544723#comment-17544723 ] 

Josef Zahner commented on NIFI-9825:
------------------------------------

[~markap14] the fact that in this case the "failure" output/loop is useless wasn't clear for me. But I fully get your point now. This is probably not clear for a lot of people. Because in my case the first impression was, that for a successful execution of the processor the flowfile comes out of the "success" output and for a failed execution it would use the "failure" output (with just attributes but no content) as usually with real incoming connections.

An UX improvement would definitely be appreciated for this use case. The introduced warning from the picture enforced this miss-understanding...

> Execution "Primary node" with incoming connections limitation bug
> -----------------------------------------------------------------
>
>                 Key: NIFI-9825
>                 URL: https://issues.apache.org/jira/browse/NIFI-9825
>             Project: Apache NiFi
>          Issue Type: Bug
>          Components: Core UI
>    Affects Versions: 1.15.3
>            Reporter: Josef Zahner
>            Priority: Major
>              Labels: ExecuteSQL, primary
>         Attachments: Screenshot 2022-03-23 at 13.53.52.png
>
>
> Somewhen in the past an information for users has been added that if a processor can have incoming connections, NiFi prevents the user to set execution to "Primary node" only. In theory this is fine, but the problem is, the "ExecuteSQL" processor can run with or without incoming connections.
> I'm using the processor on a cluster without incoming connections, but I never wanna execute the same query on all cluster nodes and it's uncomfortable to set a "GenerateFlowFile" processor with "Primary only" execution mode in front of the "ExecuteSQL". At the moment I can't set the "ExecuteSQL" to "Primary node" only without a *connected* incoming connection as NiFi generates the error message "{_}'Execution Node' is invalid because Processors with incoming connections cannot be scheduled for Primary Node only{_}". Please check my screenshot.
> NiFi should not check for the possibility on the processor for incoming connections but for real connected connections. Thanks
> !Screenshot 2022-03-23 at 13.53.52.png!



--
This message was sent by Atlassian Jira
(v8.20.7#820007)