Hey geku! We are not currently doing continuous replication between nodes, although we are working on a cloud block device backend which would allow synchronous replication using the underlying cloud block service.

In particular this would enable stateful failover... is that what you had in mind?