This article is part of our series chronicling VMware expert Mike Laverick's experience with Microsoft Hyper-V R2. He took a Microsoft virtualization course to broaden his horizons, and in this part on what he didn't like about Hyper-V R2, he talks about the lack of memory overcommit.
Memory overcommit has been a sore point between VMware and Microsoft for some time.
Memory overcommit is the VMware ESX platform's ability to allocate more memory to virtual machines than is physically available on the host itself. It allows the ESX host to achieve very high VM density, without you having to buy the physical RAM upfront. Microsoft has questioned the need for memory overcommit, despite its common usage by VMware customers. (Admittedly, I think many VMware customers are using the feature just because it is the default setting.)
Currently, Microsoft Hyper-V R2 does not have memory overcommit, but the company plans to add a new feature, Dynamic Memory, in Service Pack 1 later this year. At the moment it seems unclear, even from some Microsoft sources, what this new feature actually is. Some see it as memory overcommit, but others merely as a "balloon driver" used to recoup unused memory and return it to the host.
As far as I can tell, Dynamic Memory is not the same as memory overcommit. Instead it's a policy control that lets you set a range for the amount of memory a VM can have. However, it seems close to memory overcommit in the sense that it delivers memory on demand as the VM needs it.
Whatever the case may be, it's clear that Microsoft has taken some heat on this issue and wants to address the concern. For now, users of Hyper-V will have to buy more memory for their hosts to get the same consolidation ratios that ESX already offers. Some say this difference is meaningless, given how "cheap" memory is. This argument belies the fact that -- once you get to 4GB, 8GB or 16GB sticks of memory that are ECC- or parity-protected and branded by OEM server vendors -- RAM remains costly.
From my perspective as a VMware Certified Instructor, it's a bit flabbergasting that Microsoft is just getting around to memory overcommit, because VMware has been doing it since I started looking at ESX 2 in 2003. The fact that Hyper-V lacks this feature puts Microsoft behind VMware by seven years.
About the expert
Mike Laverick (VCP) has been involved with the VMware community since 2003. Laverick is a VMware forum moderator and member of the London VMware User Group Steering Committee. Laverick is the owner and author of the virtualization website and blog RTFM Education, where he publishes free guides and utilities aimed at VMware ESX/VirtualCenter users, and has recently joined SearchVMware.com as an Editor at Large. In 2009, Laverick received the VMware vExpert award and helped found the Irish and Scottish VMware user groups. Laverick has had books published on VMware Virtual Infrastructure 3, VMware vSphere4 and VMware Site Recovery Manager.