[
https://issues.apache.org/jira/browse/NIFI-829?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14680799#comment-14680799
]
Mark Payne commented on NIFI-829:
---------------------------------
A unit test was modified for this patch to ensure that the behavior is as
expected. However, the unit test fails due to a bug in the mock framework,
NIFI-835.
> When using InvokeHTTP, the provenance lineage indicates that it has a cycle
> and produces an error instead of showing the lineage
> --------------------------------------------------------------------------------------------------------------------------------
>
> Key: NIFI-829
> URL: https://issues.apache.org/jira/browse/NIFI-829
> Project: Apache NiFi
> Issue Type: Bug
> Components: Extensions
> Reporter: Mark Payne
> Labels: beginner, newbie
> Fix For: 0.3.0
>
>
> InvokeHTTP emits CLONE and RECEIVE provenance events but would expect a
> CONTENT_MODIFIED. This is a good example, like FetchHDFS, where we really
> need a FETCH provenance event, because the content is being modified by
> replacing it with the result of a network request.
> In the meantime, though, the RECEIVE I think should instead be a
> CONTENT_MODIFIED.
> Also, I think that a FORK may be more appropriate than a CLONE, since the
> content of the newly created FlowFile is also changed in the processor.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
