The Cloud: Chances Are You Are Doing It Wrong
By Scott Isaacs, Senior Cloud Architect and Senior Innovator, Wilmot Cancer Institute
The cloud: a magical place where all computing workflows need to reside, and if not, your organization will be left behind as your competitors move towards the future of IT. Most healthcare personnel have heard a version of this sentiment from vendors or cloud service provider representatives. The main problem with this emerging FOMO (Fear Of Missing Out) mindset driving IT decision-making is that stakeholders are asking to move workflows to the cloud to solve perceived IT issues without clearly understanding what the cloud is or what that shift means for their organization. Just because you can move a workflow to the cloud doesn’t mean you should, and many times, organizations are better off utilizing resources in their own datacenter. While at its core, the cloud is simply a collection of datacenters from which you can rent server space, in practice, it’s far more complex than that, and most healthcare IT departments are ill-equipped to navigate those complexities to successfully work in the cloud.
Currently, the cloud is in an interesting place in its lifecycle, it is both touted as part of the cutting-edge future of IT, and due to being in the market for over 10 years, it is also seen as an established technology. Like many things, the truth lies somewhere in the middle. While it is true that the cloud has been around for a while now, its growth has been similar to that of a child. In its infancy, its capabilities were very limited and its parents had high hopes that it would conquer the world. Now, it’s an awkward middle schooler who thinks he knows everything but isn’t quite ready to work yet and deliver consistent results.
Cloud adoption, like AI and other advanced technologies, is a complex journey that shouldn’t be taken lightly or as a result of FOMO.
As part of the “evolution” of the cloud, existing services are being renamed, which can lead to confusion regarding what an organization’s needs are and if they already utilize a service or not. For example, SQL Database instances were once classified as PaaS (Platform as a Service) but can now be considered DaaS (Database as a Service), which is not to be confused with Desktop as a Service (which is also known as Virtual Desktop Infrastructure). Another example is Edge computing, which once described compute nodes between datacenters and sensors or equipment, but now includes IoT (Internet of Things) devices and compute nodes that link to particular cloud providers and are deployed in a datacenter. Speaking of Edge compute nodes in a datacenter, depending on the service provider, these can also be considered a “private cloud,” which is internal hardware running software from a cloud provider, and a “private cloud” could also be a fully virtualized environment launched in an on-premise datacenter.
One of the biggest problems in the cloud is the many service providers who, in a rush to be the first to market with their platform, have not fully developed deployments for their different technologies before releasing them, and in addition, change their services quickly over time. This creates several problems for healthcare providers, including: an ever-evolving landscape of product offerings and terminology, unclear guidelines for successful deployment and integration, and similar-but-different configurations for most cloud products when compared to on-premise assets. This misunderstanding of the nature of cloud architecture and deployment, combined with a large marketing push for cloud services and the FOMO mentality, has driven many institutions to assume that, without investing in additional training and preparation, they can successfully enter the cloud space utilizing their existing IT staff and potentially a few specialized vendors.
Confused yet? You are not alone.
Even when they do recognize the need for investment, many healthcare providers do not have adequate internal capabilities, nor the budget, for cloud subject matter expertise on their team. It is detrimental to cloud adoption when organizations don’t have subject matter experts (SME’s) to perform the work, and more importantly, to vet vendors and weed out those that are just looking for a highly profitable contract at the institution’s expense. I can’t tell you how many times I’ve seen large vendors submit a proposal that covers some of the work already done by the institution and results in minimal progress in the way of full production cloud offerings, all while charging five to ten times what the market rate for the proposed work should be. Overall, the difficulties and costs of moving to and utilizing a cloud infrastructure are numerous, varied, and not to be taken lightly.
On the positive side, if you don’t know all this already, you do now.
So, now that you know the problems with cloud adoption, what can you do to better navigate these treacherous waters? As alluded to before, your best resource for a smooth and successful transition and implementation is an internal cloud SME. Nobody will look out for your organization’s best interests more than internal staff, who have a vested interest in the organization’s success. While you might be fortunate enough to find a great vendor that is more of a partner and truly does their best for your organization, vendors as industry partners typically charge more for their staff than it would cost to hire the same talent internally, and the vendor employees’ loyalty is to the vendor, not your organization. Further, the internal cloud SME’s are very likely to have a breadth of associated skills to address a multitude of IT needs. So, if you are truly looking to deploy the cloud at-scale, it is usually cheaper and better for your organization as a whole to hire your own internal cloud experts.
Another pain point at hospitals and universities is the committees put together for cloud strategies, governance, deployment, and sometimes independent departmental groups creating their own governance. While often well-intentioned, many of these committees are, in practice, never-ending meetings that only serve as a place where innovation goes to die. While it is important to get input on the overall goals and what is needed from the cloud from stakeholders across your organization (or any technology, for that matter), the discussion around purpose and use cases should happen outside of the technical processes of moving to the cloud. Governance, deployment, and final strategies should lie with the technology specialists who understand the cloud and the organization’s existing IT infrastructure. These specialists will be able to accurately gauge utilization and cost benefit compared to on premise resources, determine if purchasing new equipment or utilizing burst computing to the cloud is appropriate, or if you should deploy georedundant solutions due to having multiple locations. Separating the processes and bringing them back together for final decisions on how and when to proceed moves progress forward in a positive way. Cloud adoption, like AI and other advanced technologies, is a complex journey that shouldn’t be taken lightly or as a result of FOMO. The best strategy for implementing any technology is to have internal specialists in that field who, at a minimum, can oversee the strategy, governance, and deployment to meet the goals. Vendors should be leveraged to fill in where needed, for example, if you are unable to hire a complete team for architecting and engineering your cloud solutions, but organizations should have their own SME to oversee the vendors and take ownership once the environment is built. By resisting the fear of falling behind as well as the urge to jump into any emergent technology without properly evaluating technology staffing and business needs, your organization can avoid failed projects