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

Cloud capacity planning takes shape -- but slowly

Many companies have little to no cloud capacity planning, but as they see cost overruns in their cloud environment, they may change their minds.

Capacity planning has long been a discipline that earns lip service in IT departments, but not much more than that. That attitude has not changed much in the cloud era. It's likely, though, that IT professionals might soon devote more attention to cloud capacity planning -- at least once they realize they are overspending on cloud services.

Capacity planning emerged decades ago. At that time, IT administrators oversaw large centralized systems that cost millions of dollars. Fine-tuning the system and planning ahead for upgrades were techniques IT departments could use to minimize their expenditures.

But even then, the use of these tools was scattershot.

"Many firms do little to no capacity planning," said Jean-Pierre Garbani, a vice president at Forrester Research.

In some cases, companies lack the skills in-house. As technology has evolved, capacity planning has only become more complicated. The emergence of virtualization, for instance, means IT staff no longer examine one large system; instead, they need to track the workings of virtualized servers running tens or even hundreds of applications.

In addition, some corporations consider capacity planning not worth the investment, as costs can be substantial. Companies must purchase planning tools that have multiple components and devote staff to implementing those tools in an organization's particular environment, which requires manpower expenses. That process, depending on the complexity of an organization's IT environment, can be time-consuming and tedious.

Many firms do little to no capacity planning.

Jean-Pierre Garbani, Forrester Research

As a way around these costs and complications, a business might simply acquire more cloud-processing power when needed. In many cases, cloud vendors position their services as a cure-all for the traditional challenges that IT departments have faced when trying to deliver sufficient system resources. The suppliers keep focus on these platforms' seemingly never-ending elasticity, which makes resource allocation as simple as making a phone call or clicking a few buttons.

While allocating resources is easier in today's virtualized data center, businesses find that determining the right amount of cloud services remains a challenge.

Vendors do make it easier for IT staff to add cloud resources quickly, but like most things, convenience comes at a price. In order to be able to deliver computing resources quickly, IT departments and cloud providers overprovision resources. Of course, someone has to pay for that surplus infrastructure.

In a private cloud environment, a corporation will overbuild its IT infrastructure. A public cloud vendor, meanwhile, will have pricing models that reward customers for being steady users of a service. A customer will pay a premium when requesting a big, unexpected boost in resources.

While they tout unlimited resources, vendors face constraints on how much server processing, storage or network bandwidth they can deliver to customers. The suppliers use the airline model and overbook, so there's a risk factor that needs to be included in the planning process. If, for example, half of their customers call on the Thursday before Black Friday and ask for extra storage, the cloud vendor may not be able to fulfill all their requests.

In addition, the customer needs an adequate network connection from its site to the supplier's data center. If a vendor allocates additional capacity, it does no good if the intended users don't have enough bandwidth to take advantage of it. So, at the very least, businesses will need to continue with network capacity planning.

The cloud capacity planning tool market

Once a company decides it wants to track its cloud resource needs more closely, other roadblocks emerge. One such obstacle arises with the tools used to help determine the amount of capacity needed in particular situations: There are simply too few of those tools available.

"Established system and network management vendors have not seen capacity planning as a moneymaker," Forrester's Garbani notes.

One reason is that cloud capacity planning has become so complex. In addition to the traditional capacity bottlenecks with servers, storage and network bandwidth, new elements -- such as different cloud providers and network connections from the customer site to the cloud vendor -- have been added to the mix. Not surprisingly, potential vendors aren't enthusiastic about putting time and money into developing new products when interest in such products remains lukewarm.

That's doesn't mean there aren't planning tools available. BMC Software Inc.'s ProactiveNet Performance Management Suite analyzes, forecasts and optimizes IT resource capacity across the cloud. The system automatically generates alerts if future capacity issues emerge, so businesses can mitigate potential harm. CA Inc.'s Capacity Management suite includes what-if scenarios so businesses can examine the impact of adding various resources to their cloud applications.

Some lesser-known suppliers also offer planning tools. CiRBA Inc. claims its customers realize 40% to 70% improvements in their cloud infrastructure. TeamQuest Inc.'s products create customizable service performance reports. Uptime Software Inc.'s capacity tools enable customers to see how much capacity they are using in both physical and cloud environments.

Interest in such tools could rise in the coming years. As corporations conduct more business in the cloud, they will see the need for more effective monitoring. As they spend more on cloud services, they will want to track those expenses more closely.

Dig Deeper on Cloud automation and orchestration

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.