Breaking Myths on Platform Engineering | BizTech Magazine

Tom Henderson is principal researcher for ExtremeLabs and researches and tests enterprise-level hardware and virtual/cloud environments.

Gartner says “by 2026, 80% of large software engineering organizations will establish platform engineering teams” with the goal of providing a frictionless, self-service experience so developers can produce valuable software with as little overhead as possible. That goal is admirable, but is it realistic?

With the migration from on-premises to hybrid cloud and cloud-native apps, DevOps matured from lone-wolf team aggregation to rigorous teamwork and side trips to agile, scrum and cloud-centric approaches that were productive, if sometimes fast and loose.

In business, “go fast and break things” is a familiar refrain, but breaking things is not a good idea in software development. Even so, that ethos motivates software engineering to rise to productive challenges. Businesses experienced huge growth and transformation through DevOps and had to deal with legacy systems; legal strictures; and the mandates of federal regulators, including the Cybersecurity and Infrastructure Security Agency, the National Institute of Standards and Technology and others.

RELATED: Bake security into platform engineering.

DevOps is both a culture and a practice. Whether they have full-stack management responsibilities or partial, teams move at a strenuous pace of constant, iterative improvement. This culture creates pride, fast fixes, rapid deployment cycles and satisfaction. It’s a people- and skills-focused culture.

DevOps principles don’t ignore core issues. Development and operations teams work together, intersecting where necessary. When they don’t intersect, a theory of site reliability engineering takes hold. SRE is a monitoring function, and some have decided that it should be an additional administrative role. This has created friction between development, operations and SRE.

Conceptually, platform engineering has become a mix of operations and infrastructure. It’s not a layer on top of what software engineering entails; instead, it’s a nervous system built on much of the thinking that has made DevOps a strong success.

Platform engineering now has a track record. It’s time to separate the facts from fiction.

Click the banner below to find out the biggest benefits of platform engineering.

It can be tempting to think that if a solution works in one scenario, it will work in another; however, every platform serves a specific purpose, and what works for one might not fit another. Solutions should be tailored based on individual needs rather than a one-size-fits-all approach.

The practice requires a collective effort and thorough team and responsibility mapping to ensure that everyone works together toward the same goal. Communication must flow seamlessly among service teams, designers, security experts and other stakeholders.

Every line of code and every architectural decision represents an entry point for attack and must be viewed through a security lens. Security testing must begin from the very start of any project and continue throughout development to the product’s release.

Platform engineering can reduce costs and simplify the complexity of various tools, workflows and dependencies, particularly when developers work from an internal platform that enables scalability and hastens time to launch by incorporating reusable components such as libraries, templates and other predefined resources while also enforcing standards and governance.

LEARN MORE: Maximize your platform engineering initiatives.

A problematic adoption of platform engineering tends to increase friction unless there is buy-in from all participants and stakeholders. But platform engineering can also produce visible progress quickly, and many are surprised with successful implementations that have buy-in.

The engineering nervous system employed through platform engineering becomes the bottom layer of almost every stack. It’s an evolving definition of stack components, vetted tools and software engineering ingredients.

It’s no surprise that platform engineering sometimes becomes a rock thrown into the gears of progress. Performed correctly, it becomes part of a successful DevOps operation through thoughtful integration. But to achieve that, platform engineers must be involved in projects as equal members. Without that collaboration, platform engineering grinds slowly toward teamwork, with an interregnum of small fires and hair removal by the fistful.

DIG DEEPER: Infrastructure as code is the backbone of platform engineering.

Platform engineering techniques provide essential instrumentation for all elements of the platform, including DevOps, testing and hosted production instances, all with prescribed and consistent infrastructure. The consistency of platform engineering for the platform’s many manifestations means that there is cohesive binding between DevOps and the platform, wherever that platform exists, and on whichever turf it resides or moves toward.

This is wrong, both on-premises and in the cloud. Platform engineering isn’t a retirement plan. Instead, platform engineers are charged with keeping platforms greased for developers, software pipelines, compliance officers, security auditors, and security event and information management tools.

The resistance to platform engineering as the newest team member in software engineering often involves fear of the unknown. Many of the techniques used by platform engineers contribute to DevOps in the same way that DevOps weaves together interdisciplinary efforts between development and operations teams. Platform engineering evolves those relationships more formally, but not at the price of productivity. Instead, platform engineering can help carry the load.

UP NEXT: Five questions to ask before you start on platform engineering.