Culture / DevOps / Technology / Sponsored / Contributed

Platform Ops: The Next Frontier for Operations Teams

11 Aug 2021 9:00am, by

Karthik Krishnaswamy
Karthik is head of product marketing at NGINX, part of F5. He holds an MBA from the University of Chicago’s Booth School of Business and a computer science degree from the Illinois Institute of Technology. Originally from Chennai in southern India, he now lives in San Jose, California.

A growing number of our customers are curious about the emergence of “platform operations,” which describes teams that build, manage and optimize application infrastructure. Some are wondering how platform ops is different from DevOps, site reliability engineering (SRE), security ops and network ops. This is a valid question that requires a nuanced discussion. Platform ops is certainly the newest and least understood kid on the operations block. It absorbs some of the functions of the other ops roles while also floating above and guiding them.

Platform ops becomes necessary when an enterprise has multiple DevOps teams, each guiding its own applications and selecting its own tooling for traffic management, visibility, telemetry, security, and resiliency. In that regard, platform ops is a necessary scaling function for DevOps; this becomes even more important as DevOps capabilities and duties shift left to developers and service owners.

To put it more casually, platform ops becomes necessary when a company goes from a two-pizza team to a 10- or 20-pizza team. With one pizza, it’s pretty easy to keep a handle on what a platform is and is not. With many teams doing many things, someone has to make sure all the pizza teams get their orders, but they aren’t allowed to order from 27 different pizza joints, either.

What Is a Platform?

Before we describe what platform ops teams do, we need to pause and nail down the meaning of “platform.” The definition is necessarily fluid, but has commonalities across most companies. A platform encompasses a span of technologies touched by different engineering teams. Platform can refer to physical or virtual infrastructure and networking. A company’s platform also includes telemetry, application delivery, container deployment and orchestration with a tool such as Kubernetes, content delivery networks (CDNs) and CI/CD tooling. In the broadest sense, the platform is the environment and set of technologies upon which a company builds, deploys and delivers its applications.

Platform teams craft and curate a portfolio of technologies to maximize the efficiency of their engineering and networking teams, and establish best practices so the organization can scale more easily and securely. “But this sounds like an old-style consolidated IT organization!” some might say. Platform ops teams seek not to limit choices, but to drive consensus and ensure that everyone gets the tools and capabilities they need. This is not to say there is unlimited choice for engineers — that results in chaos. But platform ops walks the fine line of crafting and evangelizing a well-honed menu of choices that can serve 95% of needs while remaining open to feedback on the choices.

Platform Ops Through the Lens of Jobs to Be Done

What’s the best way for companies to approach platform ops? A good way to understand what a department does is to read a well-written job description for a leadership role in the department. This is the purest form of the “jobs to be done” mindset. So I thought that a great way to convey what platform ops is and does is to run through such a job description and break it down.

In this case, the job description comes from a hypergrowth early-stage startup that’s hiring dozens of engineers on many teams. This company is also building a hybrid multicloud platform on which to develop and deploy its product. The job is associate director of the company’s platform engineering organization, who will be the “…key driver of the initiatives and execution of the Platform Engineering team, working to manage and improve the team’s effectiveness, quality and efficiency.” Now, let’s look at some of the key passages in the job description that define platform ops and provide an owner’s manual on how to build a platform ops practice.

Collaborative:“…this role will collaborate across the engineering organization to create best practices and to help to build the overall ecosystem for our system and service architecture.”

Note the tone of the description. Platform ops is about collaboration and working together. This is the polar opposite of the traditional IT role, which develops enterprisewide policies and strictly enforces these policies. The tone also indicates that a platform team is a curator, a consensus builder and a coach. The most effective way to craft best practices is to do and show, not just tell. That leads to the next point.

Hands-On: “In addition to leading an engineering team, you’ll be expected to have sufficient technical skills to spend some of your time performing hands-on development work.”

Platform ops leaders and team members must be able to code, configure and deploy apps, plus all the other things that engineers do. They advise and collaborate, but they can also dive into the trenches to get stuff done. This is also, by the way, how they learn what it really feels like to use the platform they have built for the dev teams.

Innovation and Customer Service: “Above all other factors, we are looking for smart, driven candidates who want to be part of a culture of innovation and customer service.”

Innovation translates both into constantly trying out new things and continuously iterating on existing processes and products to improve. So clearly the role is not just about collaboration and curation, but also being a change agent and beta tester for novel technologies, services and products. The mention of customer service indicates that the platform team is, at its core, a type of service desk — albeit one focused on giving developers and engineers what they need to do their jobs well.

Evangelize:“Evangelize the platform across engineering organizations in order to drive continued adoption of the community-owned platform model.”

A crucial part of driving developer or engineer adoption of any new technology or product is evangelism. To be successful, the platform ops team must function like an internal developer relations center, teaching and explaining what the platform is, why it’s a good architecture and how to use it.

Establish Compliance and Standards: Establish technical standards and guidelines to ensure integrity of systems and compliance with company IT standards, policies and processes.”

Remember, platform ops is a bridge between IT and DevOps, a mashup of the two roles. Just because the platform ops team is strongly aligned with developer needs doesn’t mean that it can compromise on compliance and standards. A platform ops team must be viewed as a good team player by the IT team, not as a bunch of DevOps cowboys. This means being willing to reign in tool sprawl and make some hard choices to conform to the needs of IT policies, audits and compliance.

Work Across Team Boundaries: Establish strong working relationships at all organizational levels, across functional teams and organizational boundaries.”

Platform ops teams are not only important to IT-facing functions. Increasingly, technology is the lifeblood of every business, which makes a company’s technology platform a strategic business asset. The platform ops team provides the vascular system that delivers that lifeblood. Non-technical teams like marketing, finance and HR also depend on the platform, so the platform ops team needs to consult and build close relationships with them, just as it does with technical teams. Platform ops also needs to partner with IT; in the age of cloud and delivery of more and more software as services via HTTP, the platform is the mechanism for getting everyone in the enterprise what they want and need.

Conclusion: Platform Ops Is Coming, So Get Ready

Our most forward-thinking customers are already well into the exploratory phases of building a platform ops team. We anticipate it will be a hot topic in the second half of 2021, as the need to establish guardrails and guidelines for a “shifted-left” reality become more pressing. Even if you don’t plan to build a platform ops team, as soon as you have more than two or three DevOps teams, your organization would benefit from adopting a platform ops mindset and best practices. Fortunately, the companies that are already leaders in platform ops are happy to provide you with guidance. Look to their job listings for guidance on how to shape your approach.

Featured image via Pixabay.

A newsletter digest of the week’s most important stories & analyses.