Data Center Applications and the Cloud: What You Need to Know
It is the rare enterprise these days that does not have a presence in the cloud. And pretty soon, it will be rare for an enterprise not to have a presence on multiple clouds.
But for those who are already worrying about how to integrate relatively basic cloud architectures with legacy on-site infrastructure, the prospects of a multi-cloud future are not all bad. Between the vendor solutions hitting the channel and the industry groups developing common sets of standards and practices, it appears that unity will be easier to accomplish among cloud architectures than in traditional data center environments.
MuleSoft, for example, is staking out a strong position in the cloud integration market with its Anypoint platform designed to provide connectivity for data, applications, services and APIs across local and cloud-based environments. The system was recently augmented with the Tcat application server that extends cross-cloud integration to Tomcat development and administration systems, as well as the open source Mule enterprise service bus designed to integrate various application services. The company is banking on the hope that as cloud usage becomes more diverse, enterprises will find that commercial solutions are both easier and cheaper than custom integration coding.
At the same time, Informatica is out with a new graphical mapping toolkit that allows the enterprise to develop and track the flow of data between integrated applications. The Informatica Cloud utilizes the Vibe virtual data machine that provides layers of abstraction in order to bridge the differences between applications, many of which are launched by discrete business units with little or no regard for their ability to adapt to the broader data ecosystem. Once apps and data are dynamically linked, they can then be repurposed across the cloud for use by other business groups.
And organizations like the Open Date Center Alliance (ODCA) are working to give the enterprise community a greater say in how multi-cloud environments are to be built and operated. The group is aiming for nothing less than a fully interoperable, secure, and most of all, vendor-neutral cloud ecosystem that provides the freedom and flexibility for users to build the appropriate sets of resources without the kinds of resource and architectural restrictions that plague traditional infrastructure. The organization’s latest usage model, for example, is heavy on integration and optimization of IaaS and PaaS architectures, extending right down to services and the virtual machines that carry them.
Integration is also one of the main services that cloud brokers provide. In addition to linking users with appropriate cloud resources, a knowledgeable broker can steer clients toward integrated platforms that can extend across multi-cloud environments, even as data volumes and application complexity increase. According to CIOL.com, the brokerage industry is poised to see nearly 50 percent growth over each of the next three years as standardization among services and platforms increases across cloud providers and application developers.
Despite the attention that cloud integration is getting these days, challenges remain. The more moving parts there are in any data environment, the greater the difficulty in ensuring broad interoperability without sacrificing key operational attributes.
The safest course, then, is to incorporate interoperability as a core tenet of the enterprise cloud strategy, which will serve to foster a more coordinated data infrastructure and put the brakes on the provisioning of unauthorized, rogue cloud operations. As challenging as it may be at this stage of the game, integration will only pose more of a problem when applied to long-standing, independent cloud architectures.
