The power and relatively low cost of the cloud computing model seem like a natural fit for the education sector — but policy constraints around privacy and data sovereignty often represent a barrier to wider adoption.
But by taking advantage of a hyperconvergence model — which helps to consolidate IT resources into a single optimized platform with centralized management to enable increased utilization, boost scalability and reduce costs — can help schools earn top marks as they benefit from a web-scale, cloud-based architecture with an on-site or hosted solution.
That was the main theme of a recent ITWC-moderated webinar. Titled Education on Hyperconvergence, the session was hosted by ITWC CIO Jim Love, representatives from enterprise virtualization and storage company Nutanix (Stuart Pettifer, territory account manager, South West Ontario, and Ron Thornton, senior system engineer), along with University of Waterloo Library system administrator Adam Savage.
While cloud adoption is growing in popularity across the enterprise data infrastructure — research firm IDC notes that cloud spending will top US$500 billion by 2020 — many CIOs still maintain applications in-house for practical and policy reasons, noted Love. But the old ways of maintaining legacy systems and/or collocated IT environments can often hamstring an organization, particularly when firms can potentially benefit from running a private or hybrid cloud model. Hyperconvergence and software driven architectures promise to make those options possible using the same techniques as Google and Amazon, Love added.
The power of hyperconvergence
The enterprise challenge is server utilization — determining how much compute, network, and storage resources are needed today and three to five years down the road to avoid over-provisioning resources. This is where hyperconvergence becomes most valuable in terms of management, scalability and support, according to Pettifer. The days of x86 servers and resident storage infrastructure are becoming a thing of the past in many organizations, offers Pettifer: “Back in the day, storage wasn’t growing at the pace it is today. It’s a very dynamic world.”
Server virtualization has transformed the way servers are used in the modern data centre: the technology enables organizations to abstract compute resources (CPU and memory) from the physical servers they are hosted on. As such, the hypervisor, or virtual machine monitor (VMM) resides on one or more server hosts and virtualizes the resources, enabling users to establish single pool of compute resources for their application needs. In addition, this pool can be dynamically grown and shrunk.
With this in play, cloud and hyperconvergence can offer a more seamless IT experience. Cloud services, both public and private, means that organizations buy and use just the resources they need and only when needed. It also means they are able to “spin up” infrastructure and platform resources on demand with a single click to build, test and deploy applications.
It’s ultimately about not having to worry about the underlying infrastructure or application layer while achieving a strong ROI and rapid time to market in the process, added Thornton.
Reducing IT complexity
According to Savage, taking advantage of a hyperconverged environment addresses the University of Waterloo’s challenge of managing traditional 3-tier architecture within a large virtual server and desktop environment. As the school’s library systems administrator, Savage is responsible for managing its shared storage and backup architecture and for the reliability and effectiveness of key computing systems staff and students depend on.
The University of Waterloo library is currently running two separate Nutanix clusters — one for server virtualization and one for its virtual desktop environment. The offering is also software-defined and doesn’t rely on custom hardware for any capabilities. “One of the great things about using a hyperconverged environment is that when you need to grow, you grow in a way that suits your environment. So if you need more storage, you buy a storage-heavy node; if you need more CPU you buy a CPU heavy node,” offered Savage.
“This is a fast environment. It’s really increased the speed of our operations,” he added. The new environment is scalable, easy to manage and features pay-as-you-go affordability. It’s all about reducing complexity, he said, adding the department has gone from SAN fabric technology — and regularly experiencing lengthy firmware update times — to just clicking on a few buttons with the hyperconverged infrastructure, unlike the of multiple management panes of a traditional legacy environment.
Presently, the university is achieving a strong a 2:1 de-duplication savings ratio for the hard drive storage disk environment and an 8.41:1 dedupe savings ratio on the SSD tier, he added.
Ultimately, hyperconvergence helps to reduce complexity around infrastructure deployment and management, while helping to significantly reduce the footprint of an IT infrastructure; taking advantage of hyperconvergence means that organizations within the education sector can benefit from simplicity and fractional consumption of public cloud services while still having control over performance, data access and governance, and choice of platforms, the webinar concluded.