A Call to Action for Virtual Machine Interoperability
Jake Smith is a member of Intel’s Data Center Group focused on virtualization and cloud computing technologies. Currently working to drive standards and interoperability of cloud computing, Jake has experience in introducing new technologies in the server, storage and mobile markets.JAKE SMITH
Virtualization is a subject that is near and dear to our hearts here at Intel. We began our virtualization efforts 10 years ago, and since then we’ve worked to build platforms that enable virtual machines (VMs) to take advantage of our virtualization architecture.
To that end, we work with a very broad alliance of industry leaders to ensure that VMs operate seamlessly, and with high performance, on the Intel virtual-machine control structure (VMCS) architecture. And we’re an active member of the Distributed Management Task Force (DMTF), an industry group that creates standards that promote multi-vendor interoperability.
Given this focus, we welcome the new VM interoperability usage model from the Open Data Center Alliance, an independent consortium that works to provide a unified vision for data center and cloud infrastructure requirements. The usage model spells out actions and processes to further the development of interoperable VM management solutions.
Alliance’s Model is Positive
While it has never been Intel’s role to ensure interoperability between the VMs from different software providers, we think the Alliance’s call to action is a good one. The consortium has made it incumbent on solution providers, technology providers, and software companies to ensure that the VMs that they build going forward have a degree of interoperability.
That’s an ambitious goal. As the Alliance notes in the introduction to the usage model, VM interoperability “would greatly simplify the complexity of handling multiple cloud platforms and minimize the issues of managing workloads that are hosted on internal cloud platforms or within several different public cloud platform offerings.”
At the most basic level, the Alliance wants to see standard ways to:
- Check the interoperability of VMs
- Move or copy an existing VM from one cloud platform or provider to another cloud platform or provider
- Ensure that all operational activities can be conducted with the same syntax across different hypervisors or between two cloud providers
Those are all steps in the right direction. A basic level of VM interoperability will help cloud subscribers and providers reduce complexity and cost and improve the performance of IT services. Flexibility and choice are keys to a healthy ecosystem. It is great to see the Alliance taking a leadership position in advocating interoperability of a fundamental data center technology.
To learn more about the usage model for VM interoperability, download the PDF from the Open Data Center Alliance.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.
Nice article and very timely topic Jake:
With the various cloud, dynamic infrastructure, virtualization along with hypervisors for desktop (VDI) and servers conversations occurring in and around the IT industry, something is missing from those discussions that you bring up. What is missing is the discussion about portability of the logical or virtual machine (VM) and its underlying format such as VMDK (VMware vSphere Virtual Disk), VHD (Microsoft Hyper-V and others Virtual Hard Disk) along with Open Virtual Format (OVF) among others.
Yes as an industry we have been here before with interoperability of new technologies ranging from EBCDIC and ASCII character encoding, big and little endian machines, document formats, networks, storage and interconnects (SCSI, Fibre Channel, NAS, etc), PC vs. Apple, database formats and query languages (and programming tools), blogger vs. word press, the list goes on and on and on…
The piece that is missing from many conversations that focus on the bigger picture of being able to move VMs around private environments or between private and public clouds in a hybrid manner across different Infrastructure as a Service (IaaS) and managed service providers is that of the VM format. With many environments using VMware vSphere and thus VMDKs, that can be assumed to be the industry defacto standard. However there are also a growing number of Microsoft Hyper-V VHD as well as other formats for VMs that have been deployed for servers as well as VDI desktops and more in the wings. What this means then for portability, flexibility, mobility, agility or the common buzzwords associated with cloud, dynamic and virtualized environments is that the service provider or private (e.g. IT) entity need to support multiple hypervisor formats unless their workloads (VMs being hosted) are all homogeneous (e.g. one format).
For private environments homogenous are more likely however there is also a growing trend where IT organizations are running tiered hypervisors and thus VMs on different physical machines (PMs). This is no different than what those environments are currently doing with more than one type of network (tiered networks), different types or classes and categories of servers (tiered servers) and storage (tiered storage). With tiered hypervisors, IT organizations are realizing that they can have a heterogeneous VM environment where the right tool or technology for the task at hand can be leveraged in an economical environment. This means a mix of for example vSphere and Hyper-V or Citrix/Xen being used for different purposes that need to be managed.
Outside of the private entity or IT, the cloud and MSPs will either remain affiliated to a single VM format, or adapt to support multiple formats. That’s where there is an opportunity for open formats such as OVF to step up to the challenge of working with the large installed base of industry defacto standards to help abstract the various VM formats on a go forward basis.
In the meantime, keep the conversation going and spread the word…
For those interested, here are some related posts and links:
Whats your take on open virtualization alliance and VMware?
Server and Storage Virtualization – Life beyond Consolidation
Should Everything Be Virtualized?
Who is responsible for vendor lockin?
Chapter 10: Server Virtualization – Cloud and Virtual Data Storage Networking (CRC Press, 2011)
Chapter 12: Cloud and Solution Packages – Cloud and Virtual Data Storage Networking (CRC Press, 2011)
Chapter 13: Management and Tools – Cloud and Virtual Data Storage Networking (CRC Press, 2011)
Ok, nuf said, for now…
Independent IT Advisor, Author, Blogger, Consultant, VMware vExpert and ODCA member
http://www.storageio.com twitter @storageio