To Converge and Hyper-Converge
In This Article
For some time, the converged landscape was much simpler. Both converged infrastructure and hyper-converged infrastructure (HCI) were great choices for different use cases.
For mission-critical, tier-one applications that require guaranteed application performance, advanced data services and integration with traditional operations and management, traditional converged tends to win out. But when it comes to deploying large private clouds or supporting remote office facilities, HCI is the better option.
However, as the converged landscape continues to evolve, it's no longer a matter of converge versus hyper-converge but how both infrastructure options can work together to meet an organization's needs.
A few years ago, my industry peers and I predicted that as HCI improved, it would absorb more of an organization's general workloads; however, there would still be a converged infrastructure stack somewhere in the data center for special use cases like large in-memory databases or certain high-performance workloads such as AI/ML.
The reality is that organizations are looking for flexibility in their data center, which is why we're seeing HCI solutions like Dell EMC VxRail being modified to accept traditional storage arrays, creating these quasi-hybrid environments that are both hyper-converged and converged. The idea is to offer organizations the flexibility to absorb any type of workload in an infrastructure that is, for the most part, standardized and automated from a lifecycle management perspective. In other words, it's converged when you need it and hyper-converged when you don't.
Vendors are also simplifying their product offerings, creating momentum for pay-as-you-grow services like HPE GreenLake. Before, when organizations built complex data centers with multiple infrastructure components, it led to many disadvantages like increased operational expenses and lost time due to different requirements that needed to be met. These challenges included determining which infrastructures fit through the doors; employing more staff; and identifying how much power, cooling and rack space is necessary. To address these concerns, organizations are leveraging OPEX models that can provide what they need when they need it, reducing their upfront capital expenditures and giving IT increased flexibility.
Understanding the industry trends, how should organizations approach their data center transformations?
First, organizations must step back from focusing on specific technologies. Instead, conversations should be strategic and address what will help the business accelerate innovation. This starts by identifying your organization's business goals before developing solutions.
Additionally, be careful not to invest too much in the underlying infrastructure from an emotional perspective. Organizations often purchase technologies they already know. Some organizations will only buy from one vendor while others try to avoid vendor lock-in or only purchase white-label products because of negative experiences with vendors. It's easy to buy servers and stick them in a rack, but it doesn't necessarily transform the organization. It's about what the organization does with the infrastructure after its built. By shifting from a technology-first mentality to a holistic strategy, organizations can improve processes and achieve business outcomes faster.
In our Advanced Technology Center (ATC), customers can perform testing and validation on various converged solutions, including the solutions in our dedicated HCI lab. It's a great way to determine which converged option makes the most sense for your data center or specific use case while accelerating the decision-making process. If you're interested in exploring the ATC, contact your WWT account team to schedule time with our experts.