Making the right decisions about cloud deployments
When implementing and upgrading software, organizations must determine which solutions are compatible with their cloud strategy.
The decision to move important infrastructure into the cloud has always been daunting. Even though cloud computing has substantially matured in the last two decades, difficult choices remain.
The early criticism of the cloud was that “it’s just someone else’s computer,” underscoring the fact that the advantages of the cloud were purely logistical. A professionally managed data center would provide reliable power, networking and physical security, but the servers and the software running on them were no different than what could be hosted on premises. Unlike an on-premises server, a cloud-based server was partly outside an organization’s control, and for many CIOs, the ROI was questionable.
Determining the right level of cloud adoption is one of the challenges we have today that we didn’t have in the early days, when the choice was either/or.
As virtual machine technology came on the scene, the situation started to change. The preferred model became one in which the cloud vendor supplied and supported the environment where a VM could run, and the customer would manage the VM itself. The new model of cloud computing led to on-demand provisioning and elastic scalability, providing the potential benefits of flexibility and lower costs.
Along the way, cybersecurity became a core competency for some cloud infrastructure vendors. The cybersecurity resources of certain big cloud vendors went beyond what many of their individual customers could afford on their own.
Certifications and standards, including ISO 27001, HITRUST and Cyber Essentials, gained popularity as a way for cloud vendors to prove their cybersecurity proficiency and commit to continuous improvement.
Cloud vendors also started to develop more cloud-native services, including data storage options ranging from high-cost fast storage to low-cost slow storage. Other services, such as databases, event streaming and high availability, debuted, and pricing models changed to scale according to usage rather than the number of servers.
Today’s serverless approach
Today, it’s possible to deploy serverless solutions that use only cloud-native services with no underlying server or VM required. The new paradigm can provide better scalability and cost management.
The cloud is no longer just someone else’s computer. The “computer” is no longer literally a computer. It’s a distributed, scalable pool of computing and storage resources that can grow and shrink as needed and is priced based on consumption. And the “someone else” is a cloud hosting company with expertise at managing and securing resources.
Many software vendors now prefer cloud-native solutions because they take the variables of hardware, operating systems and site-specific policies out of the equation.
A software developer no longer needs to build, test and document their software on every possible hardware and operating system configuration and combination they want to support because with a serverless architecture, there is no hardware, operating system or other applications running on the same box. As a result, the problems of deploying, testing, distributing and upgrading their software are becoming more manageable.
Cloud deployment models
Cloud deployment models for healthcare software now include:
- Public cloud deployments, in which the vendor hosts the solution in their own cloud space.
- Private cloud deployments, which require the customer to host the solution in their own cloud environment. The customer needs to be sure that their chosen cloud vendor is compatible with the software, or else they need to be prepared to support a multi-cloud environment.
- On-premises software, which is installed on the customer’s network but likely needs to communicate with other solutions that are based in the cloud. This can require extra security and networking steps to ensure seamless deployment and operations.
- Hybrid solutions, which are deployed partly on premises and partly in the cloud.
When implementing and upgrading software, healthcare organizations need to determine which solutions are compatible with their willingness to deploy in the cloud and their ability to support cloud solutions.
Organizations also should offer space for developers to think creatively where they can develop, refine and apply innovative ideas.
Many software vendors offer a range of options to cover all four deployment models. This allows the customer to turn on just as much cloud as their strategy, policies and needs allow and require. Determining the right level of cloud adoption is one of the challenges we have today that we didn’t have in the early days, when the choice was either/or.
How do you determine the right deployment model for your organization? Consider:
- Your current workflows and environment;
- The technical requirements for your cloud environment;
- What internal capabilities are available to deliver the expected outcomes;
- Who makes up your integration developer team;
- How much time and energy your team spends on keeping applications up to date;
- How much control you want over viewing and managing connections;
- How much M&A activity have you planned and how quickly you would need to scale;
- Whether moving to the cloud makes more sense than conducting hardware refreshes;
- The amount of control you want over the configuration, development, and infrastructure management
Finding the right talent
Another challenge is finding the right talent. Today, IT organizations still have a need for traditional IT skills, including server and network management. The additional skills required in the cloud era revolve around understanding the cloud vendors’ offerings and being able to manage the organization’s software in a way that’s reliable and cost-effective.
Because so many healthcare organizations are making the shift from on-premises IT to cloud-based IT, it can be difficult to hire the right staff members to manage the new environment.
To compete, employers must offer competitive pay and benefits and a flexible work environment. They also must commit to the entire skillset lifecycle for IT team members. This includes investing in ongoing training on new technologies and providing an environment that fosters a culture of learning. Organizations also should offer space for developers to think creatively where they can develop, refine and apply innovative ideas.
Another option to consider is working with trusted interoperability partners to help reduce staff attrition gaps and minimize upskilling expenses.
Cloud computing’s journey from basic off-premises hosting to serverless, cloud-native solutions has been testing assumptions and shifting paradigms for decades. It’s time to embrace the challenges that the cloud’s evolution brings.
Drew Ivan, chief strategy officer at Lyniate, has over 20 years’ experience in healthcare IT.