We have a 7 node couchbase cluster, with one node for indexing and the remaining nodes for both query and data services. I notice that ALL the N1QL querying is occuring on one server (the master). It seems from the docs I can manually submit these queries to different nodes, but is there anyway to configure the cluster to automatically balance the query load?
How are you submitting the queries? The Couchbase SDKs should all automatically balance the load across all nodes in the cluster’s topology.
At the moment we are submitting directly to N1QL REST endpoint. We also are working on node.js client using the SDK so we will compare with that.