For the past decade, we've seen a shift from decentralized customer-owned datacenter hardware and virtualization to centralized cloud models. Although this has generally been a boost to enterprises efficiency, time-to-market, and programmability, it has shifted the location-centric workloads of the previous decade back into a datacenter-centric model where network speeds and latency between customers and the datacenter satisfy most legacy workload needs. Almost as soon as cloud developers and architects began migrating workloads into the cloud, a new requirement emerged. This requirement was driven by IoT devices, sensors, smart cities, AR/VR, and even self-driving vehicles.
This requirement was in juxtaposition to the new datacenter-centric model that scaled regionally because some of the requirements were low-latency connections, resource-constrained locations, and possibly low bandwidth or unreliable networks.
Although OpenStack has grown to be a robust, stable,...