VMTurbo drives the software-defined data center

If you’ve been faithfully following server virtualization trends over the last few years or if you’ve had your head in the clouds like me, you’ve probably already been indoctrinated into the idea behind the software-defined data center (SDDC). How pervasive has this concept become? Attend any cloud- or virtualization-related tradeshow this year, and you can bet that SDDC will be mentioned in the keynote presentation.

Software-defined data centers are considered to be the next stage of evolution for virtualization and cloud computing. It refers to a data center where all infrastructure elements — compute, storage, network, and security — are virtualized and delivered as a service. Control of that data center is fully automated and managed by intelligent, policy-driven software. But for SDDC to be successful, organizations not only have to automate as much of the data center as possible, they must also trust that process.

[ Also on InfoWorld: VMware plays up network virtualization momentum in 2014 | IBM drops $1.2B to significantly expand its global cloud footprint | Track the latest trends in virtualization in InfoWorld’s Virtualization Report newsletter. ]

How do we get there from here? The short and obvious answer is to start the journey. One possible way to begin is to use Operations Manager, a software product from a company called VMTurbo. The company describes version 4.5, the latest release of Operations Manager, as “software-driven control” for your virtual data center.

“Being able to define your data center in software is really the first step,” says Shmuel Kliger, founder and president of VMTurbo. “Knowing how to leverage all of those software levers to make sure our data centers and clouds are operating efficiently, while assuring application service levels is where VMTurbo comes in.”

VMTurbo, founded in 2009 by systems management veterans from Smarts (now EMC Ionix), takes an interesting approach to managing and controlling the virtualized data center. It essentially leverages economic principles to control the environment in a desired state where workload demand is best satisfied by the underlying infrastructure supply.

Kliger continued, “VMTurbo acts as the ‘invisible hand of the market’ that continually triggers the necessary actions to drive the data center into a healthy equilibrium and continuously keeps it in that state. Customers typically see increases in infrastructure utilization of 30 percent or more, while preventing many problems from happening in the first place.”

The company’s latest approach seems to be working, as it recently reported 2013 revenues of more than $14 million, more than doubling year over year from 2012.

With this latest release, VMTurbo has extended its market-based approach to the underlying compute fabric, with support for the Cisco UCS platform, and it’s gone deeper into the storage layer with support added for NetApp Clustered Data ONTAP. With detailed knowledge of the Cisco UCS fabric, including Fabric Interconnects, associated IO modules, blades and hosted workloads, VMTurbo now guides decisions on the right amount of compute blades and network ports to efficiently meet application demand.

Furthermore, the company has added a new “supply chain” view to provide joint operations teams with a consolidated understanding of the application workloads, their relationships to the underlying compute, fabric and storage layers and the necessary actions to move them into a healthy state.

“When we spoke to Cisco UCS customers, we found they were struggling with understanding the appropriate amount of compute and network capacity to meet workload demands,” Kliger told InfoWorld. “Consequently, many of them are drastically overprovisioning and this gets expensive fast when you consider port licensing costs.”

The Cisco UCS platform is a very sophisticated piece of equipment, forming the backbone of offerings like vBlock from VCE and FlexPod from NetApp. Fortunately, Cisco provides a comprehensive set of APIs enabling vendors like VMTurbo to build solutions on top.

According to Kliger, VMTurbo even has the capability to completely close the loop on Cisco UCS compute capacity with the “invisible hand” of the market dictating when to spin up blades on demand to maintain application performance, or spin down blades and return them to the pool when there is excess capacity.

“To adopt a true cloud operating model, we need to be able to treat our data center resources as elastic, providing the right amount of resources when and where we need them,” Kliger said.

Last year, VMTurbo took this concept a step toward the public and hybrid cloud, with support for Amazon AWS and Microsoft Azure, as VMTurbo essentially acted as a cloud broker of sorts.

As VMTurbo starts to embrace the hybrid cloud and use cases like cloud bursting, it will need to be able to make decisions around what workloads should be running where and when to assure performance while driving up efficiency, subject to policies and constraints. That means understanding when workloads can be moved to the public cloud, and when they should be moved back.

While VMTurbo doesn’t move the workloads back and forth to the public cloud today, it says it can work with other vendors to provide this capability, leveraging VMTurbo’s open APIs to help guide those decisions.

VMTurbo Operations Manager starts at $899 per socket and the new Fabric Control and Storage Control modules are available as additional licensed modules, also priced on a per socket basis.

This article, “VMTurbo drives the software-defined data center,” was originally published at InfoWorld.com. Follow the latest developments in virtualization and cloud computing at InfoWorld.com.