[
https://issues.apache.org/jira/browse/NIFI-5007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16411009#comment-16411009
]
Pierre Villard commented on NIFI-5007:
--------------------------------------
I'm not sure to understand the statement here. What I understand is: you have a
processor A scheduled to run on primary node only, this processor A has a
relationship to a processor B that is scheduled to run on all nodes. What
you're saying is that B will only run on the primary node. Is my understanding
correct?
If yes, that's the expected behavior. B will only run on primary node because A
is generating data only on the primary node, and B will be triggered only on
the node where data is available to be processed. The data between going from A
to B is not distributed accross the nodes unless you design your workflow to do
so using a Remote Process Group. You might be interested by this article:
[https://pierrevillard.com/2017/02/23/listfetch-pattern-and-remote-process-group-in-apache-nifi/]
Let me know if I can close this JIRA. Thanks.
> pre processor schedule by primary, next processor schedule by all node; but
> in fact next processor only be scheduled to primary
> -------------------------------------------------------------------------------------------------------------------------------
>
> Key: NIFI-5007
> URL: https://issues.apache.org/jira/browse/NIFI-5007
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 1.2.0
> Environment: pre processor schedule by primary, next processor
> schedule by all node; but in fact next processor only be scheduled to primary
> Reporter: Ning Sheng
> Priority: Major
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
