How can I raise debug level on tribe node?

Elastic Stack Elasticsearch
#1

Hi,
I'm facing some issues with my tribe node. Every X minutes it fails reaching _nodes thus failing the entire kibana dashboard.
I would like to debug the issue, see if its a memory issue/network. The problem is that ES tribe node logs very little information. I'm running ES version 1.7.3.

Thanks for the help!
D.

#2

https://www.elastic.co/guide/en/elasticsearch/guide/master/logging.html may be useful.

#3

Thanks, I already tried that. For some reason my tribe node doesn't accept this command:

PUT /_cluster/settings
{
"transient" : {
"logger.discovery" : "DEBUG"
}
}

Logging an error saying "error:MasterNotDiscoveredException"

#4

The error mean your tribe node is not able to talk with your downstream clusters.

Do you see all the nodes in your downstream clusters if you run the following?

GET tribenode:9200/_cat/nodes

Tin

#5

Yes, when I run

I get the list of all nodes in all my clusters.

But when I run GET tribenode:9200/_cat/master I get an empty response.

#6

Also, as described in this blog post - https://www.elastic.co/blog/tribe-node

"The Tribe node supports almost all APIs, with the exception of meta-level APIs such as the Create Index API. Meta-level APIs must be processed by the elected master node, but the Tribe node effectively has no single elected master"

Does this mean I can't raise debug level on tribe node? How can I debug it otherwise?

Thanks a lot.

#7