Hello I wrote a python program to retrieve the status of process group with this API:/flow/process-groups/{id}/status. I found that some connection status in one of nodes is exactly the same with the one in aggregated status. There are 21 nodes in cluster. I think the sum of all the nodes status equals the one in aggregated status, this is what I mean consistent status. For example, the value of bytesQueued in aggregated status should equals the sum of all the value of bytesQueued in the separate node. Could this be a bug or did I have something wrong? Any suggestions? Thanks advanced! This is the connection status in aggregated status: This is the node(bj1175) status: the connection status is the same as the one in aggregated status(last picture) This is the other node(bj1228) status: ... View more

Hello, I stopp a PublishKafka processor on the Web UI. The processor is still running in 2 hosts of the cluster. I try to find which are the 2 hosts by dragging the connection to the PublishKafka processor to another processor, I get the following message: It shows me that host bj1233 can not fulfill this request, so I check the log in host bj1233. There are two related lines: 2017-11-15 10:05:23,532 INFO [NiFi Web Server-547070] o.a.n.controller.StandardProcessorNode Stopping processor: class org.apache.nifi.processors.kafka.pubsub.PublishKafka_0_10 2017-11-15 10:05:23,532 INFO [StandardProcessScheduler Thread-1] o.a.n.c.s.TimerDrivenSchedulingAgent Stopped scheduling PublishKafka_0_10[id=6c643549-32fa-1f10-bad1-7a65935d7466] to run It makes me confused: the StandardProcessScheduler has stopped the process, but the UI shows that the processor is running. Did I misunderstand it? How to handle this situation? Any suggestions? Thanks ... View more

Hi, I'm using nifi processor(ExecuteFlumeSink) to write message to HDFS, but the messages are found in local file system instead of HDFS. The following are ExecuteFlumeSink settings: Something's configured wrong or missing? Thanks! ... View more