[
https://issues.apache.org/jira/browse/NIFI-1310?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15064827#comment-15064827
]
Matthew Clarke commented on NIFI-1310:
--------------------------------------
Not sure if this matters, but the connection feeding the RPG had back pressure
object threshold set on it.
> S2S data stops to all nodes when one node can not be reached
> ------------------------------------------------------------
>
> Key: NIFI-1310
> URL: https://issues.apache.org/jira/browse/NIFI-1310
> Project: Apache NiFi
> Issue Type: Bug
> Components: Core Framework
> Affects Versions: 0.4.0
> Reporter: Matthew Clarke
> Priority: Critical
> Fix For: 0.5.0, 0.4.1
>
>
> let me setup the scenario that causes S2S to fail.
> - I have a 3 Node cluster that is receiving data via S2S.
> - I bring down the NCM for that cluster.
> - Data continues to be delivered to nodes but sending NiFi can no longer
> update remote group peers because NCM is down.
> Everything is working as expected here.
> if one of those Nodes should then go down while the NCM is still not
> available, dataflow to remaining two nodes will stop as well within a few
> minutes.
> The expected behavior is that the sending NiFi with the RPG will fail on the
> one lost Node, but still continue sending data to the other two available
> nodes.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
