Out from under the shadow (Control-Layer: IPAPI)

Developers Blog:

A little history about Cloud Sites backend.
For many years Cloud Sites was driven by a proprietary layer on top of a Enterprise Service Bus(ESB) that maintained the veins of a complex provisioning architecture.

After years of gradual modification, the code attained a life of its own. While it was still manageable, the time required to train developers to the code and add new features became prohibitive.

We clearly needed to rewrite. But the ability to supplement the existing ESB with the customization required for Cloud Sites infrastructure was not a simple task. We also had to ensure that we wouldn’t find ourselves with the same problem as the code base matured. We implemented several future-proofing ideas to accomplish this.

– Easily consumable while maintaining secure inter-process communication
– The layer consuming the Control Plane needed to perform all the previous operations.
– Enough backwards compatibility to support the basic concepts defined by legacy schematics.
– Avoid high-cost technical debt.
– Allow quick iteration and modification by any knowledgeable team within the product.

This endeavor began in early 2015. An opportunity was presented to take what we had learned to provide the simplest rewrite available. With a multi-layer Control-Plane written in part NodeJS (Angular), Java, and then Python we began replacing the building blocks of the complex provisioning matrix. The goal was to provide our customers with a single cohesive product that manages a variety of complex components. Piece by piece the replacement came into focus.

Fast-forward to 2017 + many man-hours of dedicated sprints, we’re officially on the precipice of our second release. Finally our Legacy customers can enjoy the fruits of our labor. While some will get the immediate satisfaction of a new user-interface, behind the curtain there are major benefits:

– Faster iteration
– API driven
– Hardened security inter-process communications
– Reinforced hardware response failure resilience scenarios
– Service expansion across third party endpoints (ability to partner with external service providers)
– Exposure of internally utilized proprietary utilities to customers.

Our primary concern is that the new control-plane was built around the legacy infrastructure to ensure the safest possible transition, thus keeping our customers at the heart of this initiative.