[jira] [Commented] (NIFI-5153) When a node fails to perform a mutable request, it should be asked to rejoin cluster instead of being disconnected

["Joseph Witt (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NIFI-5153?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16463874#comment-16463874 ] 

Joseph Witt commented on NIFI-5153:
-----------------------------------

changed from bug to improvement.  Perhaps we should create a "BUGPROVEMENT" type ;)

> When a node fails to perform a mutable request, it should be asked to rejoin cluster instead of being disconnected
> ------------------------------------------------------------------------------------------------------------------
>
>                 Key: NIFI-5153
>                 URL: https://issues.apache.org/jira/browse/NIFI-5153
>             Project: Apache NiFi
>          Issue Type: Improvement
>            Reporter: Mark Payne
>            Assignee: Mark Payne
>            Priority: Major
>
> Currently, if a mutable request (POST, PUT, DELETED) is issued to a cluster, the request is replicated and if any node fails to perform the request, it is disconnected (unless all nodes fail, in which case no node is disconnected). However, often times, the failure is due to some sort of socket timeout. As a result, we should request that the node re-join the cluster, as long as its flow matches the cluster's flow.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)