Platform Engineering Improves DevOps Workflows and Results
By establishing a robust and scalable foundation, enterprises can leverage the benefits of DevOps at scale and improve the developer experience.
IN THIS ARTICLE
Challenges Around Platform Engineering
Solutions and Services to Support Platform Engineering
Achieving Success in Platform Engineering
What is Platform Engineering? An Introduction
DevOps has had a transformative influence on IT, increasing the pace and stability of development processes. Now, it’s beginning to advance platform engineering, a natural maturation of DevOps and a response to some of its shortcomings.
According to Gartner, the goal of platform engineering is “a frictionless, self-service developer experience that offers the right capabilities to enable developers and others to produce valuable software with as little overhead as possible.”
While DevOps implemented cloud-native tools that enabled scalability, availability and operability, it also represented a significant jump in complexity. That complexity has challenged many organizations, especially if their development capabilities are less mature. Other organizations have struggled to achieve desired objectives. For example, one survey found that almost 80 percent of organizations stagnate in their DevOps journeys, with limited examples of success.
DevOps is also resource-intensive, requiring customized tools and applications for each project. Platform engineering simplifies and streamlines these tools — and developers’ access to them — by creating shared, consistent resources to support developers across the organization. These may range from tools such as Docker and Kubernetes, which package cloud-driven applications in a simple, repeatable stack, to overarching practices such as Infrastructure as Code (IaC) and Policy as Code (PaC).
As a natural outgrowth of DevOps, platform engineering involves building and maintaining the infrastructure and services that facilitate efficient, cloud-native software development. Evan Bottcher, head of data and architecture at business management platform MYOB, describes these platforms in a blog post as “a foundation of self-service APIs, tools, services, knowledge and support, which are arranged as a compelling internal product. Autonomous delivery teams can make use of the platform to deliver product features at a higher pace, with reduced coordination.”
Ultimately, platform engineering helps organizations automate in alignment with specific goals so they can reap the benefits of DevOps at scale and get to market more quickly.
A Platform for Smoother Development
For developers, platform engineering resolves the tension between overreliance on operations teams for essential tasks and an overabundance of choice.
An internal developer platform helps teams become more efficient, ultimately lowering organizational costs while creating a more productive and enjoyable experience for developers.
Internal Developer Platforms treat developers as internal customers with clearly defined needs, goals and resource constraints. Self-service platforms eliminate the burdensome aspects of DevOps while ensuring that developers have the tools they need.
IDPs help to create a “golden path” that defines and supports an optimal process for fast, smooth development. That, in turn, helps to reduce security risks while ROI.
DevOps depends on continuous integration and continuous delivery, but some activities increase (CI/CD) complexity, such as changing configurations; adding variables, services and dependencies; debugging; and spinning up new environments. While platform engineering can simplify and streamline this work, it also can present challenges as organizations implement or refine engineering processes.
A PRODUCT-FOCUSED MINDSET
Platform engineering is an ongoing process, and teams must continuously improve and adapt the platform to ensure that it continues to meet evolving needs and reflect technology trends. Successful teams adopt a product management approach, conceiving platforms as customer-focused products emphasizing “value, viability and usability,” according to Forrester.
ALL SYSTEMS GO
Bringing multiple systems, technologies and applications into a cohesive platform is a complex endeavor. Such systems may include containers, configuration management tools and IaC resources, such as Terraform. In addition, the platform must communicate seamlessly and effortlessly with other systems and platforms, which means that interoperability is crucial.
ALL ABOUT PERFORMANCE
Increasing delivery speed is a top driver of platform engineering, so performance and availability are paramount. Another primary goal is to enable developers to focus on code rather than on managing infrastructure. To achieve both aims, platforms must be responsive, with low latency, fast processing and few disruptions.
SCALE AND AUTOMATE
Scalability is intrinsic to DevOps, and organizations must protect that feature when transitioning to platform engineering. As the user base grows, platforms must be able to scale up without compromising performance. Engineers also must be proficient with automation, leveraging it effectively to maintain consistency and reduce manual workloads.
SECURITY AND COMPLIANCE
Platforms must address potential security threats by implementing measures to prevent unauthorized access and protect sensitive data. Incorporating security tools and processes into platforms — and educating developers about vulnerabilities and misconfigurations — increases the likelihood of catching issues early. Platforms also must comply with relevant regulations, standards and best practices.
Learn how an expert partner can help you resolve your platform engineering challenges and reap greater benefits.
Platform Engineering on the Rise
80%
By 2026, the percentage of software engineering organizations that will use internal platform teams to provide reusable services, components and tools for application delivery1
93%
The percentage of organizations that say platform engineering is a step in the right direction; 44 percent call it a “big step”2
68%
The percentage of organizations that say platform engineering has increased development velocity2
71%
The percentage of organizations that plan to hire people with platform engineering experience within the next 12 months2
38%
The percentage of developers who use a portal or other central place to find tools and services3
93%
The percentage of top performers who use an IDP built and maintained by a platform team, using a Platform as a Product approach4
Sources: 1gartner.com, “What Is Platform Engineering?” Oct. 5, 2022; 2Puppet, “State of DevOps Report 2023,” January 2023; 3Stack Overflow, 2022 Developer Survey, June 2022; 4Humanitec, DevOps Benchmarking Study 2023, March 2023
Various motivations may lead organizations to adopt platform engineering, including a desire to increase delivery speed, scale up, reduce engineers’ workloads, control infrastructure expenses, eliminate duplicative work by product teams and simplify workflows as a product catalog grows. Ultimately, the objective is to solve problems; 54 percent of IT professionals who work with platforms say that’s the primary goal, according to a report by Puppet.
Many organizations are still assembling their platform teams and building internal capabilities. In the process, they often benefit from partnering with an expert, which can help ensure that solutions and strategies solve problems and achieve desired outcomes.
Intentionality matters: A critical insight from a recent Humanitec study is that “if you don’t build your platform, it will build itself.” While most organizations already have the major components of a platform, they also need a clear mission to guide implementation and increase the likelihood that platform engineering will meet both IT and business needs.
With that in mind, here are some important components to consider.
Scale Up DevOps Practices
The DevOps principle of CI allows for frequently merging code changes to a shared environment so that automated testing and validation can be conducted — and conflicts or discrepancies remediated — before changes advance down the pipeline. As with other fundamental components of DevOps, platform engineering can streamline and scale that practice by incorporating CI templates and other tools into IDPs.
Manage Costs
FinOps is both a practice and a set of tools that can be added to the pipeline to measure, forecast and evaluate changes based on cost. FinOps aims to improve communication among IT, engineering, business and finance stakeholders and increase clarity around trade-offs among cost, speed and quality so that organizations can monitor and manage development expenses effectively.
Focus on Developer Services
Service catalogs are essential components of IDPs. They organize and document available services across the development lifecycle, whereas IDPs contain the overarching operating tools and applications. When asked about the most critical functions of platform engineering work, teams cite service provision as a priority nearly as important as workflow automation and infrastructure building and management, according to Puppet.
Gain Visibility with GitOps
Git serves as a version control system and source of truth to ensure configuration files and source code match the desired declarations. Observability capabilities, together with continuous reconciliation between workflows and this desired end state, allow for proper tracking and approval. By facilitating this visibility, GitOps increases trust in the pipeline’s integrity.
Scale Resources Quickly
Infrastructure as Code provides infrastructure management with reusable configuration files and versioning storage. IaC helps drive fast, scalable deployment through automated resource provisioning and standardized workflows that ensure consistent and repeatable development processes. The resulting capability to scale makes this an essential tool for platform engineering teams.
Reinforce Policy with Code
Policy as Code imposes structure and implements standards set by teams and organizations using code to define, disseminate, update and enforce policy. It’s particularly beneficial for security policies, in addition to areas such as compliance, configuration management and access control. As with other types of automation, PaC reduces manual errors while increasing speed and efficiency.
Advance Security Tools
DevSecOps integrates security into development and operations processes — early and often — to discover and remediate security issues as they arise. According to GitLab, in 2023, 62 percent of developers will use artificial intelligence and machine learning to check code as part of DevSecOps, up from 51 percent in 2022.
Leverage Security as Code
Security as Code is a subset of DevSecOps that establishes a single point of security review for every change. Scans, tests and other tools allow for deeper visibility, so corrections are timely and their impact is minimized. According to Puppet, platform teams say building and enforcing security processes and improving observability are among their top goals.
Partner with a CDW Expert
Engaging a third party to help plan, design and guide a platform engineering initiative can be the best way to achieve desired outcomes, shorten the learning curve, and avoid costly delays or mistakes. Every IT environment is unique, and a partner such as CDW with experience across diverse organizations can help teams implement the right solutions and the best practices to optimize them. Key platform engineering services that CDW offers include:
• Platform Engineering Foundation
This business-orientated workshop builds a platform engineering approach to a customer’s use case, including team and responsibility mapping and a sample pipeline build.
• Infrastructure as Code Workshop
In this engagement, CDW trains a team of up to 12 people on the specifics of IaC technologies using a platform of the customer’s choice.
• DevSecOps Workshop
This workshop helps teams establish more effective security testing in an existing development pipeline, demonstrating the value of shifting security left and testing sooner.
• Multicloud GitOps Foundation
This service delivers a complete build of best-in-class offerings around Kubernetes, based on GitOps methodology and using IaC.
Part of platform engineering’s appeal is the potential to enhance nearly every aspect of the software development lifecycle.
Speed and Efficiency
Platform engineering facilitates faster time to market with a powerful combination of fundamental shifts and incremental improvements. For example, IDPs establish golden paths through pre-architected models that increase uniformity and reduce the number of tasks and decisions facing developers. Some organizations reach a middle ground of maturity (“clone and forget”) in which templates exist but are not viewed as fully developed and supported products. True golden paths serve developers effectively at every step, evolving as needed based on their feedback. Among teams that have adopted platform engineering, 68 percent report an increase in development speed, and 42 percent see “a great deal” of improvement in this area, according to Puppet. The longer organizations have platform teams, the more these and other benefits continue to grow.
Experience and Performance
Developers are the internal customer for most modern operations teams, and platform engineering helps to keep those customers happy. Self-service IDPs address the day-to-day workflows that affect developers most by significantly reducing person-to-person dependencies, waiting times, and resulting delays and frustrations. Moreover, while platform engineering directly benefits developers and their teams, it also pays dividends for infrastructure teams and entire organizations. Self-service and autonomous capabilities tie directly to development team performance. High-performing teams are far more likely than low-performing ones to allow developers to create preview environments without waiting for engineering approval or operations provisioning, according to Humanitec. High performers also have more autonomy to access information, deploy to development and staging environments, provision infrastructure, and assign resources.
Enhanced Cybersecurity and Reduced Risk
Automation, pre-architecting, consistency and stability all reduce risk. The more closely developers cleave to the golden path, the more likely they are to maintain configurations, policies and processes that strengthen cybersecurity. According to Puppet, 55 percent of teams say platform engineering improves overall security. Standardization also makes it easier to integrate the work of security teams.
Story by Neil Wylie and Peggy Alexander
