We are using a monitoring cluster to monitor multiple elastic clusters. I'm trying to write a query to check all the nodes of a given cluster name, but not sure which property in .monitoring indice would provide me with the info. I was looking at cluster_uuid, but I would prefer using a valid cluster name viz. cluster.name setting as in elasticsearch.yml file.
Any help on how to get this from .monitoring indice would be helpful.
I believe it is cluster_stats as I see the entry there, and couldn't find any entry for cluster_state type. Regardless, I'm not able to query (filter) by cluster_name as it is a dynamic field (=> using cluster_name in term/terms filter of query doesn't get me any data),
In template (.monitoring...es) dynamic is set to false though!
Below is what i see in the template.
"cluster_stats": {
"properties": {
"nodes": {
"type": "object"
},
"indices": {
"type": "object"
}
}
}
My goal is to find cluster_uuid based on input cluster name and chain it to node_stats to find all the nodes of a cluster in a watcher input.
Hi, which version of Elasticsearch and Kibana are you using? I think it was for v5.5 that we deprecated the cluster_stats type and merged its data into the cluster_state type.
Since cluster_name is not available in all data, I have to create two aggregations now and group by uuid. Say If I want to get JVM usage data by cluster/node (assuming multiple clusters are monitored) I have to get data from two places
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.