The Rise of Hyper-Converged Infrastructure

Hyper-converged infrastructure (HCI) claims to simplify IT by merging storage, computation, and typically a virtualization environment in a single system or appliance. This one-box strategy condenses the flexibility of virtualization and networked storage. According to its advocates, the end product is a versatile and high-performance system appropriate for smaller firms, branch offices, or edge applications. Furthermore, providers are increasingly attempting to supply hyper-convergence via software, particularly software-defined storage.

However, the hyperconvergence market is shifting. This is in part due to customer demand and the increasing emphasis on cloud-based and as-a-service-style consumption during the pandemic, which is expected to continue. And, just as some providers have realized that HCI is not a viable business model for them, others have joined the market, particularly with software-defined products. HCI is now well established as an on-premise alternative, particularly in deployments where administrative simplicity is critical. We have examined major developments in the hyper-converged sector and are ready to share them with you.

What’s Happening With the Market?

Emergen Research estimates that the HCI industry will be worth $48.17 billion in 2028 and will expand by 26.8% in the near future. Market drivers include those for the backup and recovery industry, as well as those for improving application performance. HCI is being utilized for VDI (desktop virtualization), databases, analytical workloads, and VM farms, according to Naveen Chhabra, a senior analyst at Forrester. And its application cases will increase soon.

Although hyper-converged is most commonly utilized in data centers, the technology is becoming more flexible as providers increase the controls available to IT managers, according to Chhabra. Nonetheless, HCI is better suited to workloads that benefit from horizontal scalability (the addition of more nodes) rather than vertical scaling (adding more CPU, storage, memory, and so on). Surprisingly, given the general market desire, some providers have backed away from hyper-convergence. 

Is HCI’s ‘Disaggregation’ Real?

The original selling point for hyper-converged was that the system’s major components are integrated. This enables IT companies to deploy technologies more rapidly while also lowering administration costs. This integration is a big part of what makes HCI appealing, especially for branch offices or edge sites. HCI, on the other hand, scales better horizontally than vertically.

Vertical scaling, such as for large-scale transactional databases, is not a strength of HCI. If your app requires vertical growth, don’t consider it hyper-converged. Although there are exceptions, monolithic systems are generally unsuitable for HCI. The reason for this is that typically, hyper-converged infrastructure cannot grow its (component) resources independently. This is leading providers to “disaggregate” or divide, the components of HCI so that they can be scaled more easily. 

Is HCI Always a Good Fit on the Edge?

HCI addresses some of the issues that organizations have when implementing technology in tiny offices, branch locations, or distant sites. These may not have separate IT staff or data rooms for larger and more complicated equipment. Similarly, hyper-converged should lend itself to edge applications, particularly when offered in a durable, appliance form factor. By eliminating the need for separate storage, computing, and networking hardware, power consumption and cooling requirements are reduced.

Using a single supplier also implies fewer moving components. It may be an exaggeration to state there is less to go wrong with hyper-converged systems – hyper-converged systems can be complicated. But IT teams should be able to oversee all of their systems from a single management console. However, no single industry definition of edge exists. Suppliers are considering a wide range of use cases, and those that claim to be able to satisfy all of them should be avoided.

HCI as a Service

The rise of hyper-converged as a service may be driven more by vendors recognizing an opportunity to deliver resources on a subscription basis than by technological advances. On one level, it makes sense for organizations purchasing infrastructure as a service (IaaS) to also acquire HCI. Scalability of cloud infrastructure on a node-by-node basis should decrease administration overheads if the workload lends itself to horizontal rather than vertical scaling. It can also make it easier to replicate on-premise HCI workloads to the cloud.

In response, CIOs must determine whether HCI is appropriate for the workloads they intend to migrate to IaaS or the public cloud in general. One advantage of the cloud is the ability to purchase compute and storage resources separately and scale them up and down as needed. Some of this intrinsic flexibility is lost when using HCI as a service. However, providers are investing in as-a-service delivery, which should make fine-tuning hyper-converged instances to varied workloads easier, especially when combined with the opportunity to utilize opex rather than capex for HCI.

Containers and HCI

Container support is one area where HCI is certainly progressing quickly. Containerized workloads are supported by established hyper-converged vendors. The ability to handle containers and hypervisors broadens the use of HCI. This compatibility is becoming increasingly crucial to customers as more cloud-based – or cloud-native – apps are designed for containers rather than VMs. Although HCI was not built for containers, vendors are customizing hyper-converged nodes to enable the flexibility required by systems. We expect HCI vendors to follow suit as more enterprise applications migrate to containers.

Final Say

Hyperconverged infrastructure has progressed from jargon to a popular concept. The technological architecture, which integrates storage, computation, and networking into a single system, has grown to the point that enterprises are seeing compelling use cases. And it has piqued the interest of larger suppliers in recent years.

“When there were start-ups a few years ago, it was all about developing and selling physical products,” explains Philip Dawson, research VP analyst at Gartner. “We referred to these as hyper-converged integrated systems (HCIS). However, as the market has shifted to hyper-converged infrastructure that overlays hardware, the architecture has shifted.”

According to a recent Evaluator Group research, business IT is increasingly accepting and using HCI. According to its survey, 79% of major companies are extending their use of hyper-converged infrastructure and adopting it for more mission-critical workloads.

“Overall, this movement represents a paradigm shift away from hardware and toward software. There was no flexibility to on-premise at the time, so it was very much a ‘you’ll get what you’re given arrangement in the early days,” Owain Williams, technical lead at Vouchercloud, agrees. “HCI is undeniably driven by larger firms making better use of their resources.”And that’s just the beginning.