Hi I have a 3 node Apache Nifi cluster setup, which is managed by a 3 node zookeeper cluster. The dev cluster worked fine, with one node frequently dropping off with us having to sometimes manually restarting the node, after renaming its flow.xml.gz and flow.json.gz, after which the node started up fine and connected to the cluster. But today, after 1 node went down, it wouldnt connect back to the cluster (even after renaming the flow gzs). Within some minutes another node disconnected from the cluster, and the last node which was the primary at that stage threw a socket time out, so I manually restarted it, and it wont startup throwing  Invalid State Cannot replicate request to Node oooo-nifiat01.yy.xxx.local:0000 because the node is not connected with the nifi-user.log complaining of  o.a.n.w.a.c.IllegalClusterStateExceptionMapper org.apache.nifi.cluster.manager.exception.IllegalClusterStateException: The Flow Controller is initializing the Data Flow.. Returning Conflict response.   It looks like the flow.xml.gz/flow.json.gz is corrupted on primary and we have a whole lot of dev which we cannot afford to lose. Could anyone please help in how we can restore the primary node, and once its online, I can bring up the other 2 nodes.    Thanks MK     ... View more