Five Questions to Ask Before Choosing a Cloud Platform Solution
A cloud of many faces is taking shape before our eyes.
It is common knowledge by now that the cloud is not merely an extension of the data center but an entirely new data ecosystem that can grow, change, and yes, evolve into a wide range of configurations, just as any organic creature is able to adapt to changing environments. In many ways, this change can be anticipated and planned for, but as the market matures, you can bet there will be a fair number of unexpected developments as well.
But to get the kind of flexibility to handle both the expected and the unexpected, enterprises will have to shed much of the thinking that has governed cloud deployments so far. This is not merely an extension of current data infrastructure but an entirely new way to doing things. Much of the attention in the cloud is turning away from simple applications and infrastructure to fully cloud-based development platforms because of this new way of thinking.
At the moment, cloud services generally fall into three broad categories: Software-, Platform- and Infrastructure-as-a-Service (SaaS, PaaS and IaaS, respectively). SaaS was the first out of the gate with companies like Salesforce hosting versions of popular enterprise applications, followed by the current surge of IaaS implementations for bulk storage and data backup. But PaaS is expected to eclipse them both in the next few years as enterprises seek to boost app development and provisioning without running up huge capital budgets. According to 451 Research, PaaS is the fastest growing cloud sector, hitting 41 percent growth in 2012 largely from the rapid uptake of Lifecycle Management-as-a-Service deployments. At this rate, cloud computing in general is expected to surge from a $5.7 billion market in 2012 to $20 billion by 2016.
PaaS’ appeal becomes clear once you take a hard look at the pressures facing CIOs these days, says CloudBees CEO Sacha Labourey. Tech executives are often tasked with providing innovative solutions to very specific problems at the drop of a hat, usually with little or no regard for the time and expense that such development usually requires. With PaaS, all the necessary test and development resources can be provisioned in a matter of hours and the app can then be launched into the cloud and then upgraded as needed with all back-end tasks running transparently in the background. If the cloud is the new data center, consider PaaS as the middleware stack.
This is the reason why PaaS is likely to emerge as the most complex component of the cloud, and the most difficult to define, according to Jaspersoft CEO Brian Gentile. PaaS gets to the very heart of the enterprise’s programming capabilities, enabling much of the functionality on which the cloud and cloud services are built. As such, it is subject to much debate as to how it should be implemented and utilized. Two key questions at this point are whether things like processes, workflows and reporting should be standardized, or are they applications on their own. Or are they both?
This programmability aspect of PaaS architectures will also drive a high level of customization in the cloud, particularly around key industry verticals. As tech blogger John Ginovsky points out, development of specialty cloud implementations like Banking as a Service (BaaS) are already underway with all promises of lower cost, operational efficiency and grand scalability that generic cloud services provide. To be sure, banks are particularly concerned about security and availability, but as the cloud becomes more commonplace those fears should dwindle. Heck, if the cloud is secure enough for the CIA, it should be secure enough for BoA.
It’s safe to say, then, that PaaS represents the future of the cloud. SaaS was a good way for the enterprise to get its feet wet with hosted applications, and IaaS provides the scalable, low-cost alternative to traditional data center infrastructure, but PaaS is where the real action is.
Once the enterprise has gotten used to distributing integrated platform environments across disparate infrastructure, the cloud should finally make the jump from a simple extension of today’s data center to a robust operating environment in its own right that is capable of forging entirely new levels of data functionality.
