Hi all!
Stack version: 7.6.0
Using the new remote monitoring (monitoring the elastic stack using metricbeat) I bumped into something that I can't find a solution for.
In the Stack Monitoring view in Kibana I now, after configuring Pipeline to Pipeline communication, always get two clusters: My actual cluster and the "Standalone Cluster".
In the Standalone Cluster I can see my single logstash node (that is using Pipeline to Pipeline communication) together with one (1) pipeline; the pipeline that is using the pipeline output. Note that the same logstash node is also included in my actual cluster, but here with only the other pipeline, the one using the elasticsearch output plugin.
My Pipelines:
input-beats.conf - Standalone Cluster
input {
beats {
port => 5044
}
}
output {
pipeline {
send_to => ["output-elasticsearch"]
}
}
output-elasticsearch.conf - Correct cluster
input {
pipeline {
address => "output-elasticsearch"
}
}
output {
elasticsearch{
# my output config
}
}
So, the data is arriving and all is well. But why isn't the first pipeline recognized as part of my actual cluster?
I'm guessing that the elasticsearch output plugin finds the cluster.uuid automatically and uses that to identify with the cluster, but isn't there any way for my input pipeline to do this as well? Or at least a way to tell it which cluster it belongs to in the configuration?
Can't seem to find an answer to this after some heavy reading on documentation and google.