Get started Bring yourself up to speed with our introductory content.

Lowering the cost of private cloud software

Don't pay more than you need to for private cloud software. These tips will help you navigate the market and implement cost-saving policies.

Establishing a private cloud builds another layer of software into the data center environment that delivers cloud functionality to enterprise users. IT administrators need to be mindful of private cloud software feature costs, as well as less-obvious ongoing or recurring costs, such as service and support contracts, which add considerably more to the total cost of private cloud deployment over time. In addition, mapping a clear cloud strategy before implementation will avoid costly and disruptive transitions to other cloud software products down the road.

The cloud software landscape

Given the number of cloud software vendors and the complexity of their offerings, evaluating and purchasing products can be a complex decision. Today's enterprises can choose from CA Technologies' suite, Cisco's Cloupia, Microsoft's Private Cloud and VMware's Private Cloud Solution, as well as a variety of open source software.

To appreciate the value of private cloud software, it's important to recognize the scope of features required to create a private cloud environment. At minimum, a private cloud must be able to inventory and pool computing resources, provision those resources to new workloads on demand and immediately scale those resources up or down as computing needs change. In addition, the self-service nature of cloud computing allows end users to do it all -- from almost anywhere using any computer or endpoint device -- without the direct intervention of IT staffers.

Private cloud software should also apply analytics to measure and track computing usage. If desired, IT can use chargeback capabilities to bill departments or business units for that usage.

It's not always helpful to compare products strictly on the basis of cost. IT professionals must look past the price tag to understand the actual value of each cloud product for the particular needs of their business.

Evaluating private cloud software

Start by evaluating the compatibility of each private cloud product. Cloud platforms should be compatible with the virtualized servers, networking and storage already deployed within the data center. Private cloud software should accommodate the data center rather than the data center accommodating a cloud platform.

Next, consider each product's feature set. Although tools carry similar core features, there is no need to pay more for unneeded functionality. For example, if the company does not intend to implement chargeback, there's no point to buying cloud software that provides it. Modular suites may allow businesses to forego unneeded modules now and save on the initial software investment, adding the modules later as needs change.

Be mindful of cost-saving deals within current vendor relationships. For example, Microsoft Hyper-V and System Center users might be able to swing more attractive pricing for Microsoft's Private Cloud software, while vSphere and vCenter users might negotiate better deals with VMware's Private Cloud. Larger enterprises will have more negotiating leverage than small or midsize companies, but it's always an angle worth exploring.

You should also consider unforeseen costs that may arise as vendor plans and business needs change. Review the vendor's product roadmap to understand how the software's features and functionality will mesh with your business plans in the future.

A business making its first foray into cloud computing might start with a private cloud deployment and not consider public or hybrid cloud integration as a meaningful feature. Forward-thinking organizations, on the other hand, might harbor hybrid cloud ambitions, so the cloud software should have that capability as a module or upgrade, or it should be on the product roadmap.

And finally, don't ignore the cost-saving potential of open source enterprise-class products such as OpenStack, OpenNebula and Eucalyptus. In a pure open source approach, code for the cloud software is freely available under a general license agreement such as General Public License or the license used for the GNU operating system. This means anyone can add to, compile and run the product without paying for its use -- and all additions or improvements belong to the project.

Open source projects are developed by the industry community rather than individual vendors, and users can modify the products for their own use. So open source is often the way to lower costs and guard against vendor lock-in, but modifications will require developers who know how to code and compile software.

You should also keep in mind that open source software is only supported by the user/developer community -- there is generally no technical support line to call with open source problems.

Enforcing policies to save money

As the business develops its cloud strategy, don't forget how proper policies and procedures can save money. Regulating resource use through monitoring, or even charging for use, can be an effective way to curb private cloud costs.

One of the biggest long-term costs in virtualization and private/hybrid cloud environments is uncontrolled, unmonitored growth in computing demand. Costly growth occurs when users, empowered by self-service provisioning, deploy and scale workloads and then eventually abandon unneeded workloads within the private cloud without scaling down or decommissioning them.

Users forget that the underlying servers, network and storage all cost the company money. As workloads proliferate, private cloud resources run short and force the business to upgrade and expand its hardware investment even though the additional resources aren't bringing any benefit to the business. The virtualization world calls this phenomenon VM sprawl, and its impact is even bigger in a cloud where users (not IT staff) tend to drive workload provisioning.

Businesses can delay and mitigate these future hardware expenses by implementing business policies that guide users and business units in the acceptable use of company cloud resources. Sharing regular resource monitoring reports is one way to raise awareness and flag users or departments that might need to curtail cloud use.

Chargeback enforces policies more directly by requiring business units to pay for cloud use from the departmental budget. Managers will often take a more proactive role in policy enforcement when it affects their bottom line.

Lifecycle management software features can also help limit resource use by reminding users of unused workloads or automatically decommissioning workloads after a prescribed expiration date.

Organizations adopt a private cloud in order to place real-time, on-demand workload management into the hands of users, freeing IT from mundane tasks to focus on more strategic and beneficial projects. But a private cloud can be a costly and challenging undertaking if you don't carefully choose private cloud software based on your current and future needs.

Dig Deeper on Server virtualization strategies and use cases

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.