IT in companies of all sizes is facing competition from the cloud. Based on a self-service, automated provisioning model, public clouds enjoy efficiencies and economics that few corporate data centers can match today. But internal IT has a distinct advantage, location. They are inside the organization and better understand what it needs. If IT can incorporate cloud-like capabilities and leverage its location advantage, they can become an efficient, cost-effective private cloud asset that propels the organization forward.
Achieving this goal, however, is going to require more than just being heavily virtualized. Virtualization is certainly an important foundation for any cloud, but it’s just the beginning. A private cloud must also provide a self-service front end and a highly automated backend. And, this journey towards a private cloud is not just for large enterprises; any IT organization that finds itself stretched thin would benefit from a self-provisioned, automated environment.
Not Just a Bunch of Virtualized Servers (JBVS)
It’s getting easier to roll out a complete compute environment. Converged and hyper-converged infrastructures, like industry leader Nutanix and VMware’s EVO:RAIL, as well as Scale Computing and NIMBOXX on the SME (Small-to-Medium Enterprise) side are enabling ‘just add water’ data centers. Vendors in this space are touting how their software-defined products can bring the cloud into these same corporate data centers. But delivering the cloud benefits to a company’s IT organization and end users requires more than simply standing up a bunch of virtualized servers.
For users private clouds often mean they have control over their own infrastructures. Self-service portals give them the ability to create VMs almost instantly and configure them exactly as needed. This translates into better productivity, as people aren’t waiting days or weeks for IT to get around to their application server requests. It also means faster development cycles and shorter time-to-market for the company. Making this work requires cloud provisioning software, such as vCloud Director, SaltStack or IBM Cloud Orchestrator, to name a few.
For the IT organization, user-managed infrastructure can lighten their administrative load and charge-back or “show-back” models common in private cloud implementations give the company better cost accounting, or at least cost realization. Private clouds can also greatly improve resource utilization by combining applications on a common server and storage infrastructure and reducing IT’s administrative burden. By pooling resources, the latest infrastructure can be made available to all VMs, allowing the most important to get shifted to newest and fastest storage and compute systems.
This can be implemented with a shared storage infrastructure, such as all-flash array SolidFire, which is fully scriptable and designed to participate in a highly automated environment. Even server-attached storage solutions, with products like PernixData, Atlantis or Maxta, to name a few, can provide enhanced performance by pooling server-side flash and RAM resources.
These are things that every IT organization strives for, reduced operational time, better performance, lower cost and improved user experiences. But what does it take to deliver these benefits? It takes more than just a bunch of virtualized servers; it takes a cloud infrastructure.
Building a private cloud requires a software layer that sits on top of that easy-to-assemble, hyper-converged architecture described above, and provides the automation and operational foundation to deliver the goodness of the cloud to users and IT. This includes automated provisioning and the delivery of compute and storage resources to users as needed via a self-service portal.
Solutions like VMware’s vCloud suite, OpenStack and CloudStack, to name a few, also take care of the back end, enabling IT to efficiently manage those resources and automate tasks like capacity expansion, performance upgrades, data protection, disaster recovery, etc. In keeping with the cloud model of efficiency, economies of scale and intelligent management, these cloud infrastructure solutions also provide analytics tools to help IT maximize performance and minimize costs.
For data centers to remain agile and be an asset to the business they must complete their shift to the cloud. Private clouds offer companies a way to leverage the cloud model of self-service efficiency and consolidation to improve end user productivity and help IT manage costs. In addition, software-defined, hyper-converged architectures have made implementing that private cloud easier than it used to be. By enabling an ‘instant data center’, companies can stand up the infrastructure they need to support their own clouds.
But virtual machines and virtualization are the first step, not the last, in this journey. A software layer is also needed to provide the operational structure for the self-provisioned, highly automated environment that allows IT to concentrate on new initiatives instead of just maintaining old ones.