Description
Request for a small change that would speed up rebalance and failover troubleshooting: could ns_server please consistently choose the same GSI node as the rebalance or failover master where possible. The same ns_server orchestrator node is already chosen consistently, but the GSI master node seems to be chosen at random. When a test or customer has done many rebalances and/or failovers, this makes troubleshooting more difficult as the timeline needs to be pieced together from the logs of many different GSI nodes.