At the last Higher Education ‘Curry Club’ evening I hosted, we got into a cracking dialogue about Public Cloud; we debated what people saw as the actual benefits, but more so we discussed the challenges of getting there and what’s holding people back.
Interestingly (to me at least), the answer wasn’t the technology…
This might seem like an obvious statement, but hosting infrastructure in a datacentre is expensive. The poor energy efficiency of most customer owned datacentres is now understood to be economically unsustainable in the long run. By reducing the amount of infrastructure needed, organisations can potentially make further cost savings, but in an age where the norm is having a converged platform; there isn’t much more to squeeze out, let alone when you run data rich services. Business cases based upon virtualisation are a thing of the past, and its forcing the question of why keep doing tech refreshes on premise?
Application environments are key to the student experience. Be it going through clearing, enrollment, or sharing exam results; students and users continually require new functionality, compliance updates, better automation, and faster processing. In an age where the iPhone has celebrated its 10th anniversary, to service these changes on traditional infrastructure is a bit old school (excuse the pun). To service modern applications and the unpredictable work patterns that drive them, traditional infrastructures are slow to expand, slower to make changes on, and typically fail when abnormal usage storms occur that exceed the platforms headroom. In other words, they’re outdated. Public Cloud, on the other hand, is engineered to consume exactly what you want, with infinite outward scalability. Take the enrollment example; thousands of students typically all need to generate and consume data within a matter of days causing a surge in activity and marking out their first experience of University IT. Public Cloud platforms scale in milliseconds when engineered and automated correctly, perform with an uptime SLA of your choice, and the second you want to stop using it the billing stops too. Consider this then the new school.
One of the biggest misconceptions about Public Cloud Hyperscalers is around cost, with it being deemed more expensive than on premises environments. For Higher Education, especially on Microsoft Azure platforms, with the right skills around designing and engineering for Cloud alongside a strong relationship with Microsoft; Cloud economics more than stack up. Better still, Cloud enables Universities to be even more effective and productive. IT teams can offload the management of infrastructure based processes (patching for example is no more with PaaS on Azure) to free up their time to be more creative and innovative.
The Journey of Cloud Adoption
Public Cloud has fundamentally different service, operational and technical architectures. Cloud engineering requires a different ethos, one of absolute efficiency and elimination of cost. Managing Cloud is a whole different beast again where failure to manage correctly has an immediate commercial impact. To get there it’s critical that a proper application analysis is conducted, where dependency mapping, availability SLAs, usage patterns and a wealth of other factors are accurately captured or defined where they previously haven’t existed. Without this, a low level migration plan is an impossibility. Hybrid identity, security and efficiency management become the lifeblood of your platform. The reality is though, the above didn’t seem to concern our guests at dinner the other evening. ANS are experts at this, with a precise portfolio of services and a long list of references who consume them.
The real challenges we debated were of changing a culture, of leading a team to a new destination and of how to truly improve services from a platform which has untold capability – and what an improved service actually means to a University.