With the demand for IT infrastructure growing at a steady rate in most organizations, embracing converged IT infrastructure allows IT operations to keep pace with demand without ramping up resources and expenses. Converged infrastructure combines compute, storage, and network in a single package, through technology such as VCE’s Vblock. EMC IT was an early adopter of Vblock in 2009/2010—first to enable our virtual desktop infrastructure (VDI) and then to support the build out of our brand new 100 percent virtualized, cloud data center.
EMC IT found that standing up IT infrastructure in pre-assembled, converged chunks offers several advantages, including a single vendor for support, a single end-of-support date for all technology components, and a single code release/support stack. In addition, due to the reduction in footprint, cabling and errors, we’ve seen faster deployment times.
Contrast this to the traditional approach to adding IT infrastructure, where you might buy storage from one vendor, compute from another and networking from a third. Each one of those individual components would have a different code set and a different end of support date. You’d be faced with managing your code independently, spending time tracking various details. What’s the microcode level on my storage array? What’s the firmware on my compute? What’s the software running on my network stack?
With converged infrastructure, IT operations can avoid grappling with such details. The vendor does all the code and compatibility testing.
Adopting converged IT infrastructure is a lot like buying a car. You don’t go to a car dealer and buy parts of the vehicle separately—the chassis, the engine, the wheels—and then assemble them yourself. You specify the features you want your car to have and then you buy the car and drive it.
With converged infrastructure, you specify what features you want from your block of combined components, and the vendor delivers the complete and tested bundle of compute, storage and network infrastructure, including the operating system tuned for virtualization. They deliver it to your data center and turn it on. Your IT operation does the finer level configurations and joins it to your network…and you’re ready to roll.
Converged infrastructure enables our IT teams to focus their efforts higher up the stack because now we don’t need to worry about turning the screwdrivers–or, continuing with the car analogy, converged infrastructure puts your IT infrastructure team in the driver’s seat instead of being stuck working under the hood.
Converged infrastructure also simplifies our planning in IT going forward. We don’t have to have someone going through the hardware vendors’ support logs to figure out if a new revision lines up with another revision. Without converged infrastructure, if I want to run a particular version of VMware, for instance, I might have to run a specific version of firmware on the compute environment, which might only support a specific version of microcode on the storage array, which would connect to a network running on a software stack that might be incompatible. It can be a tangled mess to figure out.
With converged infrastructure, however, I don’t have to worry about all that compatibility checking. I can simply set a target, such as I want to run vSphere 5.5, pull up the compatibility matrix provided by my vendor, and see the levels for meeting that target and the instructions to upgrade to those levels. The vendor, in our case VCE, has done all the planning for us.
The efficiencies are clear. As EMC demand for IT infrastructure has grown by 20 to 30 percent per year for the past five years, my team—which supports seven data centers around the world— has remained the same size. We’re letting someone else build the cars, and that’s been a big advantage in controlling resources.
Revving up the next generation
Beyond the basics of converged infrastructure, a big part of EMC IT’s infrastructure strategy is deploying the EMC Federation Enterprise Hybrid Cloud, which will definitely be built on converged infrastructure. We will actually be bringing in the next generation of Vblock™ systems for our resource pooling for this new platform, and we’ll be augmenting that platform with another type of converged infrastructure called hyper-converged infrastructure.
While regular converged infrastructure features refrigerator-size boxes of combined components, hyper-converged infrastructure features commodity, rack-mounted servers which contain components of localized storage and compute that tie to a single network space.
Our strategy is to utilize both converged infrastructure and hyper-converged infrastructure to achieve the most efficient IT infrastructure to maximize business value. This gives us the agility we need to keep pace with the rapidly growing demand for increased IT resources.
If you’re attending EMC World in Las Vegas, we invite you to attend Paul’s session titled, Enabling IaaS: Converged Infrastructure & Software Defined Storage — Tuesday, May 5th at 1:30 PM PT
Not attending EMC World? You can interact with Paul and Srini Magaluri, an EMC IT cloud architect during an upcoming CrowdChat (details here) on Monday, May 4th at 6 p.m. ET/3 PT p.m. using #AskEMCITconvinfra