Feature Request: Disable the Index Operator during a cluster rebalance

[tcrozier]

I was performing an upgrade of our Couchbase cluster and noticed that the Couchbase Operator moves and builds indexes 3 at a time while rebalancing in new index nodes. I would like it if the Index Operator would not run while the rebalance is taking place to help reduce pressure on the cluster. After the rebalance is complete then the Index Operator could process changes once again.

[brantburnett]

I am currently considering monitoring the Balanced condition on the CouchbaseCluster resource in Kubernetes (when not using an external manual-ref cluster) and disabling index management jobs while the cluster is not balanced. They could resume again when it is balanced.

Note: we are already monitoring the Available condition, this would be another required condition.