Maslow’s hierarchy of needs applied to distributed cloud data centers
There is an inexorable transition towards virtualized and software defined network architecture that use distributed data centers as vital anchor points.
These distributed data centers are going to be so fundamental to the future operation of networks that it is tempting to see what steps Abraham Maslow, the psychologist famous for his hierarchy of needs, would have laid out for this societal development.
The physiological need starts with high performance hardware infrastructure and includes compute, storage and networking resources. These will be deployed in distributed data centers close enough to users to secure the user experience for all current and future services. The major question here is the choice of technology and how far you want to be able to scale up in capacity and down in cost. The associated platform software is an integrated part of satisfying the basic data center need.
The safety need is derived from the higher dependency that society at large will have on these data centers. In the past we protected physical places and valuable things from attack. In the future we will selectively protect data and software programs located in data centers. This high level of cyber security is the largest challenge in winning the trust of businesses for large-scale enterprise application outsourcing. And the security needs to stretch all the way from the data center to individual mobile devices.
The need for love and belonging comes from the ecosystem that these data centers are a part of. Designed around open architectures, where single sources on components (x86) and software platforms (open sources) dominate, these data centers will serve an ecosystem based on open standards such as OpenStack, OpenDaylight and OPNFV, among others, as the new societal foundation.
The need for esteem is secured by all the virtualized network functions that will call these new data centers their home. Not only will they coexist with IT and public cloud applications, they will be able to do it with the well-known performance levels we know from carrier networks. In the beginning, esteem will be built with the well-defined introduction of virtual network functions from multiple vendors. However, over the long run, esteem will come from quickly introducing new functionality and performing frequent software updates at low integration costs.
The need for self-actualization stems from the fact that this infrastructure could be the first network technology to not leave a legacy network behind. With a hardware strategy in which all elements of the data center can be upgraded with optimized individual economical life cycles, there is a clear potential to phase out parts before they become a legacy. By the time we reach this level, we will likely see the whole data center complex as a platform that we can use and offer as a service.
With these fundamentals in mind, you can start to lay out a strategy for your distributed data center.