Your team uses Bigtable as the database backend. The app profile specifies single-cluster routing with manual failover when the cluster is unavailable. How can you optimize the application code for more efficient and highly available Bigtable connectivity?
