Architecting a vCloud Director Solution : Deployment Model Considerations : 3.1 Service Offerings : 3.1.4 IaaS with Disaster Recovery SLA
   
3.1.4 IaaS with Disaster Recovery SLA
None of the previous deployment options provides guaranteed disaster recovery SLAs to tenant workloads. It is the responsibility of the tenant to secure application availability by means of application-level replication. However, legacy applications do not support replication of state at the application level and must rely on replication provided by the infrastructure.
When this is the case, the provider can leverage storage replication. In the case of loss of one availability zone, the provider can restore and power on all the workloads in another availability zone.
Figure 5. DR with Storage Replication
Figure 6. DR with vSphere Metro Storage Cluster
 
There are two options for designing disaster recovery with storage replication:
• Generic synchronous or asynchronous storage replication with manual or scripted failover to the other site. For more details, see the VMware vCloud Director Infrastructure Resiliency Case Study at http://www.vmware.com/files/pdf/techpaper/VMware-vCloud-Directore-Infrastructure-resiliency-whitepaper.pdf.
• Storage solution that supports vSphere Metro Storage Cluster with automated failover provided by VMware vSphere High Availability. For more details, see the VMware vSphere Metro Storage Cluster Case Study at http://www.vmware.com/files/pdf/techpaper/vSPHR-CS-MTRO-STOR-CLSTR-USLET-102-HI-RES.pdf.