Jenkins, GitHub, or Tekton? Picking Your 2025 CI/CD Engine

Read full text Summarise with VIRAT
Authors: Daniel Rosehill, Gemini 3.1 (Flash), Chatterbox TTS
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Episode summary: The CI/CD landscape has shattered into a thousand specialized pieces. We explore why Jenkins persists as the "COBOL of DevOps," how GitHub Actions captured the default spot, and why Kubernetes-native tools like Tekton and Argo are rewriting the rules of build and deployment. From "plugin hell" to "Pipeline as Code," discover the trade-offs between maintenance overhead, platform control, and the rise of AI in the pipeline. Show Notes The state of CI/CD in 2025 is defined by fragmentation. The era of forcing every workflow into a single tool is over, replaced by a diverse ecosystem of specialized engines designed for specific environments, from cloud-native startups to enterprise-legacy mainframes. While the "just use Jenkins" era has technically passed, the old titan remains a dominant force, sitting at the heart of thousands of critical workflows. Jenkins: The COBOL of DevOps Despite being released in 2011, Jenkins retains massive market share in 2025. Its persistence is not just inertia but a technical moat built on an ecosystem of over 1,800 community-contributed plugins. This allows it to connect to virtually any system, from obscure 90s mainframes to private clouds in secure bunkers. However, this strength is also its greatest weakness. The "plugin hell" phenomenon means that a single security update can bring down an entire build fleet due to version mismatches in unmaintained plugins. Efforts to modernize Jenkins, such as the "Configuration as Code" (JCasC) movement, attempt to treat the server as a disposable container defined in YAML. While this improves reproducibility, it doesn't eliminate the fundamental challenge: Jenkins is still a stateful server requiring management of JVM heap sizes, disk space for logs, and "zombie" workers that refuse to disconnect. The Integrated Giants: GitHub Actions and GitLab CI For new projects, the default choice is often GitHub Actions or GitLab CI. GitHub Actions has seen explosive growth, driven by proximity to code; if your repository is on GitHub, the friction of setting up a separate CI server is unnecessary. It has evolved from a "good enough" microwave into a powerful tool with a secure marketplace model where each action runs in isolation, avoiding the systemic risk of Jenkins' plugin architecture. GitLab CI takes a different approach, positioning itself as a "Unified DevSecOps platform." It offers a cohesive single application for the entire lifecycle, integrating container registries, security dashboards, and built-in Kubernetes clusters. This makes it particularly attractive to high-compliance organizations like banks and healthcare providers, which require strict control over data residency and the ability to hit a "kill switch" that cloud-first models sometimes struggle to provide. Kubernetes-Native Orchestration: Tekton and Argo The most significant philosophical shift is the move toward container-first, Kubernetes-native tools like Tekton and Argo Workflows. These tools treat the cluster itself as the execution environment. Tekton, built on Kubernetes Custom Resource Definitions (CRDs), allows the cluster to natively understand pipeline concepts like "Tasks" and "Pipelines." Instead of running a Jenkins agent that executes a Docker command, you tell Kubernetes to fetch a container and execute steps directly. This enables dynamic composition and scales CI infrastructure using the same tools as production apps, such as Horizontal Pod Autoscalers and spot instances. Argo Workflows is broader, functioning as a container-native workflow engine used for machine learning, data processing, and complex deployments. Its "killer app" is Argo CD, which implements the GitOps pattern. Instead of pushing code to a server, the cluster pulls the desired state from Git and "heals" any manual deviations, representing a fundamental shift from imperative to declarative operations. Specialized and Portable Solutions The market also includes commercial powerhouses and specialized tools. CircleCI optimizes for speed to feedback with features like "Test Splitting," which automatically balances test suites across parallel containers. Bitrise has become the default for mobile development, handling the specific friction of iOS and Android builds, such as managing macOS runners and provisioning profiles. Finally, the "engine-agnostic" movement is gaining traction with tools like Dagger. Created by Solomon Hykes (creator of Docker), Dagger addresses the "it works on my machine" problem by moving pipeline logic out of proprietary YAML and into actual code (Python, Go, TypeScript). This "Pipeline as Code" approach allows the exact same script to run locally and in CI, with an intelligent caching system that avoids rebuilding unchanged parts of massive monorepos. As AI begins to creep into these pipelines, the landscape continues to evolve, offering teams more choice—and more complexity—than ever before. Listen online: https://myweirdprompts.com/episode/2025-ci-cd-tool-landscape

Fields of Study & Subjects

Software Engineering Techniques and Practices Software System Performance and Reliability Software Engineering Research

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
30 March 2026
Language
EN