For a replicated bigtable table with two clusters one in region X and one in region Y, I assume requests near region X go to region X nodes and requests near region Y to go to r