If there is no acceptance for data loss or service interruption (note2) you are most likely looking at Synchronous replication with Metro Cluster from NetApp or something up that alley , this is typically followed with a high hardware cost but if the requirement is there so is the funding.
Data Replication is a small part of disaster recovery planning (note3) , replicating VM’s can help speed up the recovery time so instead of having to restore from tape or the local d2d device , so just adding Hyper-V replica to the party does not solve everything
Replication can also be handled at the application , so Exchange and SQL both have ways of helping out with ensuring data
Hyper-V Replica gives everyone else than the top tier a out of the box method to enable replication on all or selected Virtual Machines hosted on Hyper-V V3.0
Hyper-V replica don’t have any requirements other than a IP stack and a Hyper-V server to replicate to , so we can have a shiny SAN on the primary location and a stack if 7200’s on the 2nd location for replication target , this will still give possibility to start up critical VM’s until we run out of IO.
But in these times it can be hard to get funding for IO that sits around doing nothing so a replication to slow drives beats no replication any day.
(note1) that can live with data loss from the last sync to the point in time where the primary storage system went down.
(note2) there is always a risk
(note3) read http://workinghardinit.wordpress.com/2013/03/30/a-reality-check-on-disaster-recovery-business-continuity/ , and repeat after me data replication is a very minor part of disaster recovery