2012-02-22

Cloud waves

Of all the cloud-related topics making headlines these days, hybrid cloud is arguably the one that is attracting the most interest from IT leaders. But they are stymied because of the relative immaturity of the tools and frameworks that support hybrid cloud. The necessary tools and frameworks will therefore mature in three waves over the next few years.



The first wave, which we are experiencing now, centers on point-to-point, mostly batch-style, data integration between enterprise applications and software-asa-service (SaaS) applications. The early adopters of this wave include enterprises that need to expand into a new geography and opt for a SaaS-hosted model for some applications, requiring data integration with existing applications. Other early adopters include those that need new functionality added to an existing system, which requires integration with external data sources in the cloud—social networks, for example. Another set of early adopters: those that need analytics applications where external sources are ingested or internal sources are published to cloud-based analytics systems.Today, there is no shortage of tools for point-to-point integration, but frameworks for data integration that support hybrid cloud are few and far between, especially those that help with large-scale cross-cloud master data management. Furthermore, consistent data virtualization is still difficult to put together in a hybrid cloud context.

The second wave will see workflow that spans on-premise and off-premise services where the business process involves both. This will demand integration at data, service, and business process levels, potentially across multiple clouds. The challenges look much the same as in the first wave, but with far greater complexity, and a higher degree of consistency is required at multiple levels. Cross-cloud security management, for example, becomes critical here.

The third wave could be described as “white label” integration of enterprise applications in a hybrid cloud model, seamlessly mixing on-premise and offpremise services in the package. There’s still a long way to go before cloudbased software development lifecycle tools and application frameworks have built-in support for hybrid cloud. End-to-end management across all aspects of the applications and systems is virtually nonexistent.

Extract taken from Accenture's  Technology vision 2012

No comments:

Post a Comment