Why Developers Shun Ops: Rethinking DevOps and the Rise of Platform Engineering

Introduction

Full text is about 5,500 characters, requiring roughly 7 minutes to read.

The original title considered "DevOps is dead, platform engineering is the future" but was changed to a more provocative wording; the article re‑examines DevOps and platform engineering, discussing concepts and core ideas for internal development platform (IDP) practitioners.

Goals of DevOps

Since its 2009 inception, DevOps emphasized team collaboration, automation tools, and process improvement to accelerate software development and deployment. After nearly 15 years, many organizations still face high delivery costs and deployment friction that go beyond tooling.

DevOps is a philosophy that encourages close cooperation between development and operations, but it does not prescribe a concrete roadmap, leading to the emergence of platform engineering. A July 2022 tweet "DevOps is dead, long live Platform Engineering" sparked widespread discussion.

Platform engineering is a newer operations concept that stresses self‑service capabilities for developers. By abstracting infrastructure complexity, it provides flexible toolchains and workflows, enabling teams to solve problems independently and improve delivery speed and quality.

Platform engineering is the discipline of designing and building toolchains and workflows that provide self‑service in the cloud‑native era. Internal platform products (IDPs) cover the entire application lifecycle. – definition from platformengineering.org

Both DevOps and platform engineering aim to improve software delivery efficiency and quality, but platform engineering focuses on reusable platform architectures, scenario‑driven capabilities, and a self‑service experience, whereas DevOps emphasizes collaboration, automation, and process improvement.

In 2023 Gartner listed platform engineering as a top strategic trend, and it appears again in the 2024 technology trends, indicating growing industry recognition.

Recent high‑profile incidents (e.g., Alibaba Cloud outage on 11‑12 and Didi outage on 11‑27) highlight the critical importance of operational stability and the need for robust platform practices.

Why Developers Reluctant to Do Operations

• Focus on core development tasks: Developers prioritize coding and have limited time for operational work.

• Lack of interest or expertise: Many are unfamiliar with or uninterested in ops responsibilities.

• Heavy operational burden: Ops involves 24/7 monitoring, high‑impact incidents, and on‑call duties, creating additional pressure.

• Insufficient tooling and platform support: Poor automation forces manual effort, raising cost and complexity.

Nature of Operations Work

Operations ensure system safety and stability, requiring continuous monitoring, resource management, incident handling, and routine maintenance.

11‑12 Alibaba Cloud incident: a logic flaw in AK code caused a global outage lasting three and a half hours, resolved by batch restarts.

11‑27 Didi incident: a low‑cost K8s in‑place upgrade was performed incorrectly, causing a downgrade that killed all pods; service was restored only on the following afternoon.

These incidents serve as stark reminders of the challenges all enterprises face in maintaining online stability.

Some Reflections

Safety first, always: Speed should never compromise operational safety.

Shared responsibility: Both developers and ops personnel can cause costly failures; everyone must be vigilant.

Stability depends more on meticulous execution of countless details than on any single technology, and measuring success remains difficult.

Platform engineering was introduced to address these pain points by providing self‑service, scenario‑driven capabilities that streamline software delivery.

Key Success Factors of Platform Engineering

How to Promote Platform Engineering Inside a Company

• Platform scope: Establish authoritative tools, invest continuously, and avoid duplicated efforts.

• Platform culture: Serve developers as primary customers while satisfying management needs.

• Platform goals: Build scenario‑driven, self‑service tools.

• Platform owners: Assign clear owners for each scenario to eliminate ambiguous responsibilities.

• Demand source: Prioritize developer needs and avoid overly broad version upgrades that increase migration cost.

• Platform APIs: Provide rich, well‑documented APIs for internal consumption.

From a global perspective, these points outline how to drive platform engineering adoption.

What Kind of Tools Should Be Built Under Platform Engineering

• Productization: Internal tools must be treated as products serving the entire organization, not just a single team.

• User experience: Offer intuitive GUIs and APIs while shielding complex backend logic.

• Integration: Combine monitoring, logging, alerts, and other observability views into a single workflow for each scenario.

• Self‑service: Enable developers to accomplish tasks without platform team assistance, akin to using an ATM instead of a teller.

Common issues for internal development teams include productization drift, high‑priority ad‑hoc requests, lack of value recognition, and duplicate building.

Recommended actions: continuously collect user demand, maintain comprehensive documentation, provide open APIs, actively promote the platform, and collaborate to avoid siloed tool development.

How to Measure Platform Engineering Success

Key metrics can be grouped into user, product, and organization dimensions:

User dimension: weekly active users, number of enabled business units, NPS.

Product dimension: access efficiency, execution efficiency, success rate.

Organization dimension: cycle time, effort hours, etc.

These indicators help assess the value delivered after a year of operation.

Future Outlook

International Situation

Leading companies such as Google, Spotify, Netflix, and Walmart are actively adopting platform engineering. In November, CNCF released a Platform Engineering Capability Maturity Model, evaluating teams across five dimensions and four maturity levels.

Domestic Situation

In China, platform engineering is gaining attention, especially among those previously responsible for DevOps tools. The China Academy of Information and Communications is drafting a national standard for platform engineering capabilities, and I am participating in this effort.

Gartner predicts that by 2026, 80% of software engineering organizations will have platform teams, with 75% offering developer self‑service portals.

Platform engineering is not just a trend; it is the future of software delivery.

Recommended Operations Tools

Resource Management (Taishan CMDB)

Provides unified resource management, tagging, multi‑dimensional queries, and standard APIs to support automation and intelligent operations.

Operations Foundation (Taishan Tiger Symbol)

Unified host agent management platform offering Open‑API integration for monitoring, chaos engineering, bastion, and other core services.

Host Security (Taishan Bastion)

Enterprise‑wide bastion platform with comprehensive permission control and security audit capabilities.

Operations Automation (Taishan Qilin)

Automation platform enabling self‑service scripting, command execution, and workflow orchestration to improve efficiency and reduce human error.

Cloud Management (Taishan Cloud Governance)

Unified management of public and private cloud resources, delivering self‑service capabilities and simplifying cloud resource usage.