Featured Articles

Place Resources Toward Most Important Cloud Services

By Casey Houser
Contributing Writer

Even the most successful and domineering enterprises in their respective fields have limited budgets. They must prioritize where they place resources, and if officials handle such prioritization poorly, over time it can send things crashing down.

A recent blog post from David Linthicum at GigaOm suggests that business officials organize their cloud services into three tiers: business-critical applications, business non-critical applications, and occasional-use applications. These three tiers can help provide officials with rough estimates of how much money and other resources they should dedicate to their cloud applications.

The rubric can work well for any sort of hardware, software, personnel, marketing, or other issue. Cloud services, though, are becoming among the most essential parts of modern enterprises, and the relatively new adoption of some services may suggest that they are more important than they seem. In fact, that might not always be the case because occasional-use applications, for instance, will not demand the same amount of uptime or die-hard security as mission-critical applications.

Business-critical applications require high availability and high elasticity, Linthicum says. These are the applications that run inventory control, sales management, and logistics systems. If they stop working, they could cost businesses millions of dollars each day, and as such, they require specific service-level agreements that demand high availability, fault tolerance, and scalable performance. The price of these important services is not always in line with their importance, so they do not always demand a pretty penny. They are not, however, services on which businesses will want to skimp.

Business non-critical applications demand medium availability and low elasticity. Though they are not useless, they will not cause businesses to stop functioning if they stop working. They demand SLAs that are less strict and costs that are less expensive than applications in the first tier. Linthicum notes that, in his experience, cloud services tend to have high uptimes regardless of their costs. Again, businesses will get what they pay for, but these applications will not demand high costs as much as they will demand legal protections through their SLAs.

Lastly, businesses will require occasional-use applications only for tasks such as quarterly inventory. These may include online storage services that businesses can use for free or at very low costs. They are mainly necessary to support business efficiencies. Their performance is not as critical as any application in the first two tiers, and officials may classify them as demanding medium availability and no elasticity.

Officials that use a rubric such as this one can fit in their existing and future applications, and alter budgets and resources accordingly. If they are placing too many of their eggs in one basket, they can back off and reallocate resources to applications which demand more than officials are presently giving them. It should help maximize the value of cloud services businesses use and make for clear policies that will dictate future actions.