Arthur Cole spoke with Rob Peglar, senior fellow, Xiotech. Virtualization has moved from a cutting-edge enterprise technology to an integral component of data center infrastructure in record time. So it's only natural that organizations will start experimenting with multiple virtualization platforms, which adds yet another facet to already complex management and automation tiers. A new breed of management firm is relishing this challenge, devising new ways to bridge the divide between hypervisors. Xiotech is one, as Peglar explains.
Cole: More virtualization management firms are building multi-hypervisor capabilities into their platforms. What are the main challenges in overseeing disparate virtual environments?
Peglar: There are two main challenges. One is the human aspect of learning, sufficiently and robustly, to manage these environments. The other is the intelligent use of storage in multi-hypervisor environments. Hypervisors that can communicate directly with storage, through techniques such as REST, will overcome the challenge in an optimal fashion.
"There are two main challenges. One is the human aspect of learning, sufficiently and robustly, to manage these environments. The other is the intelligent use of storage in multi-hypervisor environments."
Cole: There must be operational benefits to, say, running VMware and Hyper-V side-by-side. Do users see any benefit, or is it strictly on the managerial side?
Peglar: Users will see benefits in the form of business optimization. There is always a natural tendency to move toward the hypervisor, which is easier, faster and cheaper to operate and use. But as well, the business can benefit operationally by moving workloads and applications back and forth between the environments to take advantage of the inherent efficiencies in each.
Cole: What exactly is your Intelligent Storage Element (ISE), and how does it facilitate management in multi-platform environments?
Peglar: The Intelligent Storage Element can be thought of as a storage blade, much in the sense that data centers use blade servers to consolidate their compute assets into scalable and manageable entities, all the while running multiple platforms, such as operating systems, hypervisors and applications, on them. The ISE facilitates management by its design of independence. Each blade has its own resources, including cache, capacity, bandwidth, power/cooling control, and is high-performance and high-reliability as well. So the trust in the storage blade is present from a business point of view yet all the blades are managed as a single collection -- again much as a collection of blade servers is managed. The element-based approach has proven to be optimal in multi-platform compute environments, and it is the same with the ISE for storage environments. Key to this management is the use of standard protocols and techniques such as REST and Web services, facilitating optimal multi-platform conformance and interoperability.