DevOps platform engineering -- the next generation of DevOps
A decade ago, DevOps emerged as a cultural phenomenon, bringing developers and operations professionals together to break down silos. However, as extraordinary of a shift in IT this was, DevOps has quickly matured to the point where discussing it as a practice seems outdated.
Regardless, the transition from the original concept of DevOps to its current state signifies its triumph rather than its demise. In fact, from the ashes of DevOps rises the platform engineer, a new role specifically tailored for the modern era of DevOps. Under the umbrella of platform engineering, DevOps now operates with a dedicated budget, a team, and a suite of self-service tools that empower developers to directly manage operations.
The platform engineering team offers significant advantages, enabling developers to leverage Kubernetes as a self-service tool. This enhancement boosts the efficiency and speed of development for numerous users, signifying the maturity and ubiquity of Kubernetes. According to Gartner, within the next three years, four out of five software engineering organizations will leverage platform teams to provide reusable services and tools for application delivery.
Platform engineering serves as the new middleware in the software development landscape. As the number of developers and applications continues to soar, the traditional concept of middleware -- an always-available app server -- has been replaced by platform engineering's self-service model for developers.
This transition holds immense significance. During the earlier phases of DevOps, experimentation with various technologies was widespread, but these technologies had not yet converged. However, modern applications now operate seamlessly using containers and storage, with networking and security managed through Kubernetes in a cloud-native manner.
Developers no longer rely on ticketing systems. Instead, they expect elastic infrastructure that they can use and deploy through the platform, which is maintained and operated by the platform engineer. The platform engineer ensures that the platform is self service, highly available, reliable, elastic, is multi-tenant yet highly secure, has guard rails to prevent overextending the platform, provides usage monitoring and bill back methods but the platform engineer themselves are not building the application or own the deployment of the applications. They are enablers for speeding developer innovations and application deployment. This shift in maturity improves responsiveness, allowing developers to make swift changes to their applications and rapidly drive them into production. With developers taking charge, both development and deployment times have drastically decreased.
T-Mobile, for instance, reduced application deployment time from six months to just hours with the assistance of Portworx as an element of their platform architecture. Enterprises with thousands of developers require self-service or on-demand access to storage and data services, which platform engineering teams strive to deliver at scale.
As a progression from traditional IT, the platform engineering group relies on two key sets of technologies: cloud-native technologies and AI-ready data services: modern databases and data services like Postgres, Redis, Cassandra, Kafka, and Spark, and more recently analytics and AI tools like Snowflake and ChatGPT -- all provided as services by the platform team to developers.
Platform engineers offer crucial services that eliminate the need for extensive Kubernetes expertise among users. These services include managing various Kubernetes distributions like OpenShift, GKE, EKS, or Rancher, as well as providing security measures through platforms like Prisma Cloud or Sysdig. Additionally, platform engineers handle data on Kubernetes, managing storage resources, backup, disaster recovery, databases and data services under the Kubernetes umbrella. We are witnessing the efficiencies firsthand, with a small number of platform engineers catering to hundreds of users.
When a technology becomes ubiquitous, it gradually fades into the background. Consider semiconductors as an example -- they are present everywhere, powering our remote controls, phones, and cars. However, as end-users, we rarely consider their existence, they have become invisible. Kubernetes is experiencing a similar transformation. In the enterprise realm, Kubernetes is becoming deeply ingrained in various systems, and the self-service paradigm makes it inconspicuous to users. In the past, every developer needed to possess comprehensive knowledge of Kubernetes. Now, developers only need to utilize it, leaving the intricacies to the platform engineer. Platform engineering will make Kubernetes invisible.
Platform engineering bestows a valuable gift upon developers by relieving them of the burden of comprehending Kubernetes at a granular level as part of their daily responsibilities. This eliminates a skills gap issue as Kubernetes continues to flourish, and is relied upon by all of the world’s largest and most successful organizations. Platform engineers and Kubernetes are a match made in heaven to aid a company’s innovation and competitive edge.
Image Credit: Sergey Nivens / Shutterstock
Murli Thirumale is Vice President and General Manager, Cloud Native, Pure Storage.