we did an updgrade from ES 2.2.1 -> ES 2.3.1, Kibana 4.4.2 -> Kibana 4.5.0 and using the latest Marvel version. Agent is installed on each node (incl. license) and after start of Kibana we only get data from the master node. Same after switching master to different node:
This is the result of the new node resolver feature not picking up the transport address properly. You can workaround this issue by changing the node resolver to use the node's name as the unique properly rather than the transport address (the default). This is done in the Kibana config (kibana.yml):
marvel.node_resolver: name
Given that you do not have have domain names in your transport address, can I have you run this to ensure that the forthcoming fix is enough:
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.