Infrastructure as a service: Why it’s more viable than ever
Organizations need infrastructure that can scale quickly to adapt to rapid growth and changes in the market. Flexible consumption models are delivering on their promise.
  • Paul Gillin
  • 11/05/2021
Infrastructure as a service: Why it’s more viable than ever
IT agility is more important than ever. Organizations need infrastructure that can scale quickly to adapt to rapid growth or changing customer demands. In the wake of COVID-19, the months-long procurement and installation cycles that have been the norm in data centers for years are no longer acceptable. Moving everything to the cloud, however, isn’t an option for most mature organizations. Migration is time-consuming, expensive, and may not even be an option for workloads that contain sensitive or regulated data. IT organizations also have existing investments in facilities and equipment that are not easily abandoned. While many businesses intend to move more of their processing operations to cloud platforms over time, they want to do so at their own pace.

New options are emerging that borrow from the “as-a-service” software model and bridge to hardware infrastructure, from data center systems to endpoint devices. This approach lets the customer maintain full control over how and where infrastructure is deployed without taking ownership of it. Instead, the service provider assumes responsibility for installation, operations, maintenance, and upgrades, while charging the customers only for the infrastructure and software they use.

There’s a lot for overburdened IT organizations to like in this cloud-like consumption model.

“Many of our clients spend 60% of their time keeping the lights on,” says Dale Aultman, Vice President and General Manager of International Infrastructure Solutions Group Services at Lenovo. “If we’re running the system and the middleware for them, they can focus more on the business and reduce day-to-day operations.”

The infrastructure-as-a-service market is poised for rapid growth. IDC estimates that worldwide annual recurring revenues from dedicated (local) cloud infrastructure-as-a-service offerings for compute and storage will increase from $138 million in 2020 to $14 billion in 2025, with a compound annual growth rate of 152%.

Part of the reason is that organizations are learning a lot more about the true value and cost of public cloud services. While cloud pay-as-you-go pricing is attractive in the short term, the reality is cloud platforms are often more costly over three- to five-year time period, particularly when zombie instances and data egress fees are considered.

One Lenovo customer that was looking to migrate to the SAP S/4HANA enterprise resource planning application determined that hosting the software and the cloud was three times as expensive as deploying it locally. “It helps to engage with someone who can help you analyze what is best suited for the public cloud versus on-prem or hosted,” Aultman said. 

Best of both worlds?

As-a-service hardware and other infrastructure services from a third-party provider can be the best of both worlds in such scenarios. Customers offload the overhead of managing servers, storage, and networks to a third party with significant expertise in that area. Capital expenses are reduced, and costs shift to a more predictable operating expense stream. The service provider is contractually bound to ensure that sufficient capacity is available as well as to handle infrastructure additions and upgrades transparently. Lenovo, for example, typically runs its clients’ infrastructure at 40% capacity to ensure that there is always plenty of room to expand when needed, Aultman said.

This arrangement may remind some IT veterans of data center outsourcing, a trend that flourished in the 1990s and is still used by some enterprises today. However, infrastructure-as-a-service is different in important ways. Outsourcing firms typically bought equipment and hired staff from their clients, then sold that same infrastructure back to them as a service for an annual fee. Clients had little control over how operations were deployed, and upcharges buried in contractual fine print often sent costs skyrocketing. Customers also had little control over who was running their infrastructure because outsourcing firms redeployed the acquired staff to meet their own needs.

In contrast, the new breed of infrastructure services is predicated on a long-term alliance between service provider and customer, with each committed to the other’s success. There are no surprises and no loss of control. CIOs should look for a provider that can work with them to analyze the characteristics of each workload and choose infrastructure options accordingly. Factors to consider include how much data movement is involved, the importance of preserving existing infrastructure investments, regulatory compliance requirements, and the organization’s most pressing modernization needs, Aultman says.

It’s becoming clear that not only is the as-a-service model for IT infrastructure viable, but it may soon become the standard for how organizations deploy on-premises systems and endpoint devices. “Everything can be provided as a service,” Aultman says. And soon, perhaps, it will be.

Watch the latest episodes of LNIT

Season 2 (8 episodes)
Season 1 (8 episodes)