Industry focus on stack providers such as VMware and the various OpenStack implementations has prevented operators from focusing on the core problem: optimization of the underlying resources.
Leveraging an integrated Policy-Driven Model-Based Service-Orchestration and Resource-Automation framework, we developed a solution where we can optimize the underlying resources, and where the span of control for any given virtualization stack is what is dynamically managed. This helps us address several of our critical use cases in addition to optimizing raw underlying resources including: (a) power management - turning off un-utilized resources, (b) services of bare-metal requirements for CDN, some real-time processing applications, etc., and (c) managing the needs of specific virtualization stacks (including isolation into secure network zones) required for specific VNF and/or IT domains.