Follow

Ingestion in NiFi Cluster

Issue

When ingesting data into an environment with a clustered NiFi set up and multiple Zookeeper instances, an error my by encountered that involves the inability of NiFi to communicate within each machine.  The symptom gives the appearance that one NiFi machine stays up and stable and the other will seems to start and then stop.

Cause

Multiple Zookeeper instances are listed in the NiFi.properties file when only one should be.  Also related to Interset issue number https://filetrek.atlassian.net/browse/FT-15679

Resolution Step

On each NiFi machine open the nifi.properties file.

Look for the line nifi.zookeeper.connect.string

In cases where this is an issue, there will be a list of Zookeepr nodes.  This needs to be a single instance of Zookeeper and should be the same one for each NiFi node.

For example.  If the file appears as such when opened:

nifi.zookeeper.connect.string = machine:2181, machine2:2181

Change to the following format on each NiFi machine. Please ensure that the same value is present on each machine running NiFi.

nifi.zookeeper.connect.string = machine1:2181

Applies To

  • Interset 5.6 or higher 
Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request

Comments