If you read my book, you'd know that I advocate for the position that service oriented architecture (SOA) enables an enterprise to be adaptive -- by committing to the agile approach to service-enabling business processes, an organization exposes IT infrastructure with shared/reusable web services.
In my experience, a big benefit of cloud computing (for example, the cloud model for the U.S. Department of Health and Human Services' implementation of the Federal health care exchange, as I envisioned while on task at MITRE) is how that approach minimizes capital investment, keeping costs proportional to traffic/usage. The additional benefit of ubiquitous services (including automatic hot site / failover, and load balancing across geographically dispersed data centers) means purveyors of cloud infrastructure provide a large component of the tools needed to achieve, well, the Zen of SOA.
Think about this scenario: using Amazon or Google, an enterprise may provision and deploy onto the cloud infrastructure new or additional instances of both a platform (CentOS, for example) and software on an as-needed basis. As new users subscribe/access, the cloud platform addresses increased provisioning needs. In the case of AWS or Google, this underlying platform expands quickly and cheaply. Oh, and also contracts when not needed.
When I've developed policies and procedures for a service-enable enterprise, I've witnessed how that governance can provide controls to manage web service life cycles. With the SOA approach (read about it here), CIOs can track modifications to services, and place controls over who can change or update such services. With my kaizen-modeled philosophy, this occurs in a repeatable fashion, when governance policies are enforced. From this vantage point, a service-enabled enterprise can leverage the benefits of a cloud model, almost from the get-go.
No comments:
Post a Comment