Expert Xu Lei on DevOps Efficiency, Practices, and Toolchains

There is no enterprise unsuitable for DevOps; every company that seeks profit can benefit from increased efficiency, and therefore should adopt DevOps.

Guest Profile : Xu Lei, founder and chief architect of LEANSOFT, focuses on software engineering and DevOps consulting, with over ten years of R&D project management experience, former GM of SSW China R&D Center, Microsoft MVP, certified ScrumMaster and agile coach.

Understanding DevOps : Xu defines the core competency of DevOps as "efficiency". Over the past decade he has been improving software delivery speed, noting that DevOps does not solve business problems directly but enables faster, higher‑quality delivery of business requirements.

He traces the evolution from SDLC to ALM, agile, and lean, explaining that DevOps returns to concrete technical practice while ALM focuses more on management.

In practice, DevOps improves team morale, accelerates requirement grooming (e.g., reducing a two‑month process to two weeks), and automates repetitive tasks such as iOS build and signing, turning manual steps into continuous integration pipelines.

Suitable Enterprises & Evaluation : All enterprises can adopt DevOps, but success depends on people. Xu cites a bank that achieved full‑lifecycle management only after reorganizing into a virtual cross‑functional team, highlighting the importance of cultural and structural change.

Key Implementation Points :

Top‑down cultural shift

Bottom‑up practice support

Engineering execution that bridges the two

Leadership must recognize DevOps value; without it, bottom‑up efforts will fail. Practices such as Scrum, Kanban, CI/CD provide the necessary support.

Career Path : Xu emphasizes learning from failures, noting that early DevOps work was often misunderstood and undervalued, but perseverance leads to expertise.

Current State & Future Trends : DevOps is still emerging; while management methods and toolchains are mature, successful case studies are scarce, especially in traditional IT enterprises. The main trend is adapting DevOps to complex legacy environments.

Tool Recommendations :

Lifecycle Management Platforms :

Microsoft Visual Studio Team Services / Team Foundation Server (requirements, project, configuration, test, build, release, analytics)

Atlassian suite: Jira, BitBucket, Confluence, Bamboo

Automation Engines :

Jenkins – widely used, extensible, plugin‑rich

TeamCity – enterprise‑grade, strong unit‑test reporting

Code‑Quality Tools :

SonarQube – multi‑language static analysis platform

Coverity – static analysis for C/C++/C# with security focus

Parasoft dottest – coverage and runtime testing

Roslyn – .NET compiler platform for dynamic analysis

FxCop/StyleCop – Visual Studio static checks

CheckStyle/FindBugs/PMD – Java static analysis

Automated Testing Tools :

Unit test frameworks: JUnit, NUnit, Google Test, xUnit, Mocha, Jasmine

Functional testing: Selenium, Appium

Performance testing: JMeter, LoadRunner, Visual Studio Load Test

Environment & Application Orchestration :

Docker – container platform (including LinuxKit, Docker Swarm)

Kubernetes, Apache Mesos, Service Fabric – cluster and application orchestration

Chef, Puppet, PowerShell DSC – configuration management

Virtualization & Cloud Platforms :

Public clouds: Microsoft Azure, AWS, Alibaba Cloud, QingCloud

Private clouds: Azure Stack, System Center, OpenStack, VMware, KVM

Xu stresses that no single tool can cover all DevOps needs; organizations must build a tailored toolchain that fits their processes and culture.

Training Course : The upcoming "DevOps Practice" course combines theory, real cases, and hands‑on labs with a full online environment, treating each session as a real project to give participants practical DevOps experience.