A PureCloud Platform Story
Call me Bob
I’m a CTO. What keeps me up at night? Worrying about a lot of things: competitors with better tools, the endless cycle of upgrades and equipment investments, and disasters we don’t see coming.
I worry about change… The cycles of technology are accelerating, but picking the wrong technology can kill a company. I know we have to keep moving forward. I’ve been following cloud development for years now, and what most people don’t realize is that it’s not just a bigger and better data center—it’s a truly new way of building software and services.
The real question is not whether we’re moving to cloud services; it’s how long can we put off moving everything there!
Built to scale, built to change
To build great cloud services, you have to start over from the beginning. Porting legacy server software just moves old bottlenecks to a different system. Big, complex processes have dependencies that don’t scale. We’ve been working around this for years with object-oriented approaches, but with old-school internal software dependencies and wait states, a failure in any one process can cause a devastating cascade of problems.
From monolithic applications to microservices
I paid attention when Google introduced a new approach. Instead of using big, complex processes that loaded a whole database to do queries, they moved to small, lightweight processes that could only do simple tasks, but could be massively duplicated on pools of servers to attack huge problems. Instead of locking out other processes to complete a query, they “hold on loosely” to data and take the best current answer available.
Software architects call this approach to building small parallel processes “microservices.” PureCloud nails it. It takes time to build software like this, but PureCloud has been at it since 2012.
No more updates, thanks to continuous deployment
It’s sad, but even when companies test their software rigorously, enterprise software updates can cause headaches. When they add new features, they break your old requests. And even if everything goes well there are unexpected side effects like bigger data storage needs. Modern software is extremely complex, and even with test automation and Test Driven Development (TDD) there are things that get missed because no other company has your unique combination of enterprise systems and hardware and customizations.
A platform for innovation
PureCloud takes a different approach. They build and deploy many times a day, with no interruptions to their customers. New changes are automatically tested with increasing rigor as they promote through multiple deployment environments before reaching production. A core feature of the platform are the PureCloud APIs which provide a contract for integrating with PureCloud services. The APIs maintain backwards compatibility as new features are added, so I rest easy knowing our existing integrations won’t ever break.
PureCloud keeps working even if the internet is down
Unexpected things can occur… Hard drives fail, connections are lost, servers lose power and a thousand other things can go awry. PureCloud Active-Active architecture utilizes multiple availability zones, so there’s never a single point of failure. They further mitigate risk by using automation to introduce random failures and ensure services recover. I sleep better knowing that when things do go wrong, I’m using a system that’s built for resiliency and recovery.