[ 
https://issues.apache.org/jira/browse/NIFI-632?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14580480#comment-14580480
 ] 

Matt Gilman commented on NIFI-632:
----------------------------------

I haven't looked at the patch but just some background on what I think is 
happening with the annotations.

The annotation invocation is based on my comment above [1]. Since the 
processors are being instantiated to extract properties/relationships, they are 
given the chance to do some initialization logic. In one of my cases, this 
involved setting up an ExecutorService for running a background thread. My 
processor expects to be given the opportunity to shut this down as part of the 
normal processor life cycle. This was missing initially. 

[1] 
https://issues.apache.org/jira/browse/NIFI-632?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14573597#comment-14573597

> Documentation Generation: Nar ClassLoader
> -----------------------------------------
>
>                 Key: NIFI-632
>                 URL: https://issues.apache.org/jira/browse/NIFI-632
>             Project: Apache NiFi
>          Issue Type: Bug
>          Components: Core Framework
>    Affects Versions: 0.1.0
>            Reporter: Matt Gilman
>            Assignee: Dan Bress
>            Priority: Minor
>             Fix For: 0.2.0
>
>         Attachments: 0001-Squashed-commit-of-the-following.patch
>
>
> When auto generating component documentation the NarThreadContextClassLoader 
> needs to be set in case the component attempts to load anything off of its 
> classpath. Without this, the documentation generation fails.



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

Reply via email to