Microsoft has frequently carried statements about supportability of Exchange Server in a hardware virtualization environment (see Microsoft support policies and recommendations for servers that are running Exchange Server in hardware virtualization environments), and Exchange Server 2010 documentation contains details about exactly what Exchange 2010 configurations are supported in virtualized environments (see Exchange 2010 System Requirements). Frequently the Exchange team blog has also carried posts about virtualizing Exchange. But customers continue to ask questions.
Microsoft answers your virtualization questions once again in a blog post on the Exchange team blog, and also addresses the misleading guidance from VMWare about deploying VMWare’s high availability features in vSphere along with Database Availability Groups (DAGs), Exchange 2010’s own high availability feature.
Of course, if you’ve asked this question in Exchange communities, you would’ve received a prompt reply back from Exchange MVPs, Microsoft employees or other community members — mixing two high availability features (one that’s application-aware, Microsoft’s DAG; and one that works on a lower layer and isn’t application-aware) doesn’t make sense, and adds needless complexity to a deployment. Not to mention the additional costs of deploying an additional high availability technology.
To read the official Microsoft response, head over to Answering Exchange Virtualization Questions and Addressing Misleading VMware Guidance on the Exchange team blog.