What Is DevOps? Benefits, Challenges, and Balancing Efficiency with Stability

At the 2017 Operations/DevOps Online Technology Summit, senior expert Lian Ming from Alibaba Cloud shared insights on DevOps. The article starts by defining DevOps, then compares it with traditional models, lists its difficulties and problems—such as finding a balance point, responsibility division, and assessment constraints—and finally provides a concise summary.

What Is DevOps?

DevOps is an engineering model that maximizes efficiency by clearly dividing responsibilities among development, operations, testing, and pipeline roles. It focuses on role division rather than organizational restructuring; a vertical hierarchy does not guarantee the required division, nor does a horizontal structure imply traditional division.

The goal of DevOps is to achieve maximum engineering efficiency. It is a methodology designed to reach that goal.

DevOps vs. Traditional Model

In the traditional division, product designers (PD) propose requirements, developers write code, SCM packages it, QA tests it, and operations (OPS) deploys it, culminating in a release.

Advantages: clear division and responsibilities, guaranteed quality, layered controls, and easy management.

Disadvantages: high communication and waiting costs, each stage can become a bottleneck, and OPS often ends up as a “clean‑up” role dealing with bugs.

DevOps Division Model

In the DevOps model, developers drive the entire process through tools—automatic packaging, testing, deployment, upgrades, and monitoring. SCM, OPS, and QA act as guardians around the tools, ensuring each step functions correctly. The tools must support evolving business changes, and when they cannot, manual intervention is required.

Benefits: reduced communication and waiting risks, shorter delivery cycles, and developers own the delivery, avoiding hand‑off disputes.

Drawbacks: many participants increase risk, high cost to support diverse business scenarios with tools, manual fallback needed when tools fail, slower professional growth of tools compared to humans, and potential over‑empowerment of developers.

Challenges and Issues to Solve

Finding a Balance Point

DevOps aims for maximum efficiency, but efficiency and stability often conflict. How can we improve efficiency while ensuring stability?

In traditional models, OPS handles stability; in DevOps, developers assume this role, creating tension when a single team must meet both business and stability KPIs.

Responsibility Division

Developers focus on coding; packaging, testing, and release are secondary tasks using tools. Determining how much non‑coding effort developers should allocate is crucial for smooth tool usage and business agility.

Tools act as the glue, needing openness for other teams to optimize each stage and must evolve sustainably.

Constraints and Assessment

After breaking the old balance, a new one must be built. Increased developer voice brings new constraints, both internal and market‑driven. Companies must find a balance, define responsibility, and establish assessment mechanisms to sustain the model.

How to Measure DevOps

Engineering efficiency: time from requirement receipt to production release.

Stability: high efficiency without stability leads to rapid failure.

Non‑development work ratio: a large proportion indicates risk of failure.

Business scale vs. ops staff ratio: e.g., Google SREs manage ~2,000 machines each.

Summary

Automation may reduce the need for traditional ops staff, which is an inevitable trend.

DevOps has no single best practice; focus on case context and business background.

There is no absolute good or bad between DevOps and traditional models—only suitability.