A. Remaining members recover by repartitioning the data across the remainig cluster members in parallel with normal request processing
B. Some latencies may be experienced due to higher priority of recovery
C. In flight operations may be lost
D. A system administrator must manually assist recovery of the lost data

A. Guarantee of higher application performance and reliability if the transactions are flowing between instances of the same container type.
B.Easier to build a Shared Services infrastructure that can provide best practices， tools and expertise around HA/Management.... with a single type of container than a mix of multiple types
C. You’ll save on hardware costs with a homogenous environment
D. It’s a necessary step in our move toward adopting SOA