I have a 5 node elasticsearch cluster where all the settings are the same on all nodes. Marvel does not see one of the nodes and in elasticsearch.log I can see these lines repeating again and again:
[2015-05-07 02:44:13,814][ERROR][marvel.agent.exporter ] [ES_FOUR] error connecting to [[0:0:0:0:0:0:0:0]:9200] [Connection refused]
[2015-05-07 02:44:13,814][ERROR][marvel.agent.exporter ] [ES_FOUR] could not connect to any configured elasticsearch instances: [[0:0:0:0:0:0:0:0]:9200]
Nodes are all in the same network, settings are the same. Any ideas what can be the cause and what to check to troubleshoot this?
Thanks in advance.
I had exactly the same issue. Apparently, this is related to marvel trying to use IPv6 for exporting data. If IPv6 is not used and not properly configured, the simplest way to solve this is to force marvel to export data using IPv4 address.
Add this to your elasticsearch.yml (on all nodes):
marvel.agent.exporter.es.hosts: ["127.0.0.1:9200"]
and then restart Elasticsearch on each node one by one. (Perform rolling restart)