Cloud-First Strategies Call For New IT Operations Models

The trend is clear. Many enterprises are adopting a cloud-first approach. Gartner, Inc. says that by 2024, more than 45% of IT spending on infrastructure services will shift from traditional data center solutions to the cloud. This implies that C-level executives need to figure out how to evolve their IT operations.

There are many operational models from which to choose, including CloudOps, FinOps, SaaSOps, DevSecOps, NoOps, AIOps — and the list goes on. That can make it difficult to know what you need. However, as you embrace hybrid and multicloud strategies, having the right operating model is crucial.

Familiarize yourself with the shift to software-defined infrastructure.

Adopting new models is important because, in hybrid and multicloud environments, there is a shift toward software-defined infrastructure. Evaluate whether you can benefit from site reliability engineering (SRE), which is basically approaching IT operations as a software problem.

Google pioneered SRE for operating its services and explained that adopting SRE implies “continually defining reliability goals, measuring those goals, and working to improve [its] services as needed.” There is also an emphasis on breaking down silos between development and operations teams and resources having an understanding of the entire application lifecycle.

Rethink skills and job roles.

Hybrid and multicloud environments call for multifaceted talent. Adopt more multiskilled resources as opposed to the siloed resources that are common in traditional IT operations.

For example, you would traditionally have roles like a network engineer or a compute engineer. In today’s software-defined multicloud environment, however, you need to approach things more broadly. For instance, you might have a cloud engineer who operates a whole slew of cloud infrastructure services across multiple providers.

Understand that other shifts also call for new operational models and multiskilled resources. There’s a shift from infrastructure as a service (IaaS) to platform as a service (PaaS) and software as a service (SaaS). Meanwhile, lines are blurring between applications and infrastructure, as there is increased adoption of serverless computing and container technologies. Now, there’s no strict delineation between an infrastructure resource and an application resource. The growing use of AI-led automation for self-healing systems, infrastructure as code and application deployment as code is also contributing to this shift.

Understand that IT evolution is a journey — evolve and encourage collaboration.

You don’t have to select and stick with just one operating model. Be aware that you might need a combination and that you need to continually evolve — and understand that collaboration is going to be key.

Initially, you may start off with adopting IaaS on cloud and embrace the CloudOps model. In this mode, you will still be responsible for cloud infrastructure operations like monitoring and patching. As you start consuming more SaaS applications, your operating model evolves to be SaaSOps with no infrastructure operational responsibility.

SaaS offerings are more turnkey and generally less expensive in terms of total cost of ownership. However, some SaaS providers might make you believe that they have everything covered, so there’s a tendency to underestimate the level of IT operations you need to adopt SaaS. Be aware that you’ll need your own IT operations model to address seemingly simple things, such as identity management for SaaS applications and integration with employee lifecycle management.

Continue collaborating and evolving your models as you go. To establish an effective cloud FinOps model, for example, you’ll need cross-organization collaboration. Bring together organizations, including business, finance, HR and sales, to figure out how you can optimize your cloud deployments and expand your opportunities for savings. Tapping different disciplines should enable you to understand the whole story.

Grow your talent via training.

The in-house staff providing your current data center support functions and, more broadly, your IT operations staff might be prime candidates for cloud training and certification. Provide these individuals with opportunities to get the skills to make your new operating models work.

If you have the resources, consider establishing an in-house cloud training program to provide your associates with the needed coaching, certification and hands-on experience. If establishing a program is not an option, identify existing training and certification university programs offered by established, respected cloud experts. Enroll your most promising people in these cloud programs to build the teams of cloud engineers and other cloud-certified folks you need.

Traditional IT operations models may have been useful in the past, but you need new models to succeed with a cloud-first strategy. As Unisys research shows, one in three cloud migrations fails to realize the benefits of a cloud transformation.

To ensure cloud transformation success, adopt new models, collaborate and develop new skills.