Embracing the Self-Service Cloud

Data Center Applications and the Cloud: What You Need to Know

Of all the benefits that cloud computing brings to knowledge workers, none is more intriguing than self service.

Sure, scalability, flexibility and lower costs are great for the enterprise, but if you are a data consumer, nothing compares to the feeling that you can get whatever you want, whenever you want it, with no questions asked other than “What is your credit card number?”

For DevOps teams in particular, self-service is one of the primary drivers of both public and private cloud infrastructure, say Red Hat’s Malcolm Herbert and Roger Nunn. Under the old regime, resource provisioning was slow, complicated and rarely met project requirements. This, in turn, led to diminished productivity and a failure to truly leverage the enormous investment that the enterprise had made in increasingly cumbersome and complex data infrastructure. Self service (and, this being Red Hat, open source-based self service) allows resources to be brought up in minutes and lends support for a fully automated DevOps process in which developers can concentrate on actual development, rather than environment building and maintenance.

Small wonder, then, that the cloud industry has placed self service as one of its top attributes. Portfolio management specialist RightScale, for example, recently launched RightScale Self-Service, a user portal that provides instant access to cloud infrastructure for a wide variety of applications and services. Not only does it provide a single avenue to needed resources, but it also offers centralized control and governance to the enterprise, which should go a long way toward controlling the shadow IT that arises when users take it upon themselves to build their own environments and fill them with company data.

Self service does have its drawbacks, however. Some enterprise applications require highly specialized operating environments and resource configurations, and it’s not always easy for the user to tell what will work ahead of time. That’s why providers like Microsoft are tying self-service capabilities to key partners and consultants who can guide the user to the proper deployment. In a way, this is much like the traditional vendor/reseller paradigm in which not only the software product is delivered, but the means to deploy and integrate it into legacy architectures as well. By next month, Microsoft is expected to roll out the Open Licensing program for the Azure Cloud, which should simplify the relationship between enterprise and developer even further.

Ultimately, all of this should coalesce around the hybrid cloud, where self service, cloud bursting and a range of other capabilities combine to form a unified environment that is both easily manageable and highly functional. But as KEMP Technologies’ Jason Dover explains, this won’t come about by itself. IT executives will have to plan for and build a host of advanced capabilities, such as real-time capacity planning, data location and dynamic networking to ensure that data and applications can find each other and then deliver productive, timely information to the end user. The tools needed for all these capabilities are available—the only challenge now is to implement them in a cohesive fashion.

Self-service provisioning will definitely be a step up for enterprise productivity, but it can be a double-edged sword if not architected properly. In particular, the enterprise should take care to include provisions for the reclamation of resources once they are no longer needed, preferably using processes that function as quickly and easily as the initial deployment. As well, careful attention should be paid toward potential conflicts—at some point, even virtual resources have to draw from the physical world.

Nonetheless, a world in which individuals are better able to achieve their own goals without a lot of centralized hand-holding can only be to the collective benefit of the enterprise.