Comment by jdc

5 years ago

Isn't this a solved problem? Just use a failover, right?

If your consistency model allows you to use their asynch replication and potentially lose data in surprise switchover - sure. Otherwise you’re looking at similarly worse performance and also worse reliability