Application Reads on which Cluster

Team,

Looking for a right metric to determine which cluster is taking read traffic. We have 2 clusters for redundancy at anytime only one cluster will take read traffic anc both takes write traffic.

I have been playing with metrics like query_current from _nodes/stats/indices/search , _nodes/stats/thread_pool?pretty=true | jq -r '.nodes | . | .thread_pool.search.active'.
All these gives me similar data. Is there a metric that clearly states reads or application read queries are routed in clusterA or B.

Hello @mannoj87

As per the documentation

Can we try to use below :

GET /_nodes/stats/indices/search?filter_path=cluster_name,nodes.*.indices.search.*

query_current
query_total

Thanks!!

Yes tried it, in both clusters I see numbers like

cluster1 : query_current=9

cluster2 : query_current=10

The value is sum of all data and query nodes falling under {indices.search.query_current}.