4.) Disaster Recovery SLAs – If you are considering Cloud for Disaster Recovery, having a documented SLA in hand is absolutely critical. I think everyone will remember the outage Amazon suffered in 2011, The point is that something bad happened, some lost data, and many claimed Cloud was not ready for prime time. Further there were some who were blaming the customers for not architecting for failure and knowing what is and is not covered in the SLA. Still others blamed Amazon for not delivering.