Taking Cloud Platform Services Seriously

(Perspective from early 2009)

Based on conversations with colleagues and clients, I still don't think that the majority of long term, enterprise architecture strategies even consider Cloud Platform Services (referring to PAAS)...but I think this strategy may eventually become mainstream.

Despite the fact that the cost of hosting systems on-premises is reducing with time due to improved virtualization and management technologies, I think that in the not too distant future, there will be a point when the costs and risks of serving up certain types of applications in house will actually be greater than any compromise made by moving the same systems to a PAAS, cloud based platform.

The biggest drawbacks I see with PAAS are:
1. Perceived loss of security controls at the infrastructure level.
2. Perceived loss of control over infrastructure configurations.
3. Vendor lock-in.
4. Vendor price gouging.
5. WAN/Bandwidth limitations.

Despite the risks, I do think that visionaries are beginning to consider PAAS for certain aspects of their computing environment road map. I'm just not sure how much longer it will take before this approach is generally accepted and considered mainstream.