您的浏览器禁用了JavaScript(一种计算机语言,用以实现您与网页的交互),请解除该禁用,或者联系我们。 [思博伦]:持续测试在成功实现CI/CD中的关键意义 - 发现报告

持续测试在成功实现CI/CD中的关键意义

信息技术 2022-11-04 思博伦 LLLL
报告封面

The Criticality of ContinuousTest for CI/CD Success The Criticality of ContinuousTest for CI/CD Success Introduction The benefits ofCI/CD (continuous integration and continuous delivery[or deployment]) through DevOps adoption have been increasinglysought in today’s fast-evolving technology landscape for the agility itoffers development and the stability it provides to operations throughthe automation of the technology lifecycle. A growing number ofleaders view CI/CD as the solution to help maintain their competitiveedge. Too often overlooked in the big picture, however, is the importanceof continuous testing (CT) to make all that possible. The compellingbusiness benefits that flow from getting it right are significant:heightened productivity, accelerated speed to market, improvedquality and substantial cost savings. The fundamental significance ofeffective CT and methods of fostering its success in an organization’soverall CI/CD strategy is the focus of this white paper. Components include: •An overview of the evolution of DevOps and testing automation•Common challenges for realizing reliable CT•Adopting the right strategic approach to achieve CT•Benefits of a mature lab as a service, test as a service andDevOps adoption•Solution partner selection criteria for achieving comprehensive CT The Evolution of DevOps and Testing Automation 15 years ago, the organizational focus in telecommunications was on optimizing productivity in development andproduction. No one was really looking at the testing process to see how it could or should be improved. As the industryevolved over the years, however, every component of the telecom innovation pipeline came under scrutiny to identifyareas for improvement. Test automation gradually became a core requirement. Then, acknowledging the complexityof testing required in the emerging digital marketplace, a host of new challenges were revealed. Testing leadersbegan to ask: How do you automate builds and make them faster? How can we provide fast and continuous feedbackregarding the latest pipeline build or release candidate? The advent of DevOps.Around ten years ago,organizations that automated and incorporated theseenhanced processes produced decisive accelerationfor development and operations. These innovativeprocesses, along with the advent of Agile practices,and the gradual phasing out of waterfall model insoftware development, contributed to the emergenceof the strategic approach which became known asDevOps, with CI/CD its fundamental objective. The emergence of Continuous Integration (CI). This quest for test process improvement fosteredthe development of distributed compilation withthe creation of compiler/build farms that deliveredcontinuous compilation to identify issues early andoften. These process enhancements ultimatelyevolved into continuous integration (CI). CI in turnbecame adopted on an increasingly wider scalewhich influenced the way organizations approachedproduct development. These beneficial applicationswere ultimately applied to the process of building anddelivering a product or piece of software within anorganization, or out to a customer. A closer look at Continuous Testing (CT).At the coreof ensuring CI and CD, CT involves the execution ofautomated tests that optimize the software deliverypipeline. A continuous test process is empowered by: •Many short cycles•Aggregated changes•A large pool of repeatable tests•An orchestrated pipeline•An elastic infrastructure Common Challenges for Realizing Reliable CT While the goal may be simply stated, understandingthe ‘what’ and ‘why’ of obstructive factors that canhinder the prospects of achieving reliable CT, earlyon in the planning phase, is essential. Gating factorsof success in CT exist in three areas; all influence eachother continuously: These inefficiencies in test design, siloed testmanagement and misaligned priorities in testing, resultin needless and time-consuming corrective work thatcripples the advantages of CT. The end result is thattrue CI/CD is impossible. Process — It’s a marathon not a sprint.While somesolution advocates expect results once a new solutiontechnology has been integrated into an organization,that’s a misplaced expectation. Recognizing thatdisparate teams must converge in a unified processis no simple proposition. Uniform best practices mustbe adopted across all groups, instilling productivecollaboration, to assure all parties are working fromthe same criteria and requirements. As well, a group familiar with Agile, DevOps and leanmanufacturing principlesis in a different position ofsolution adoption maturity than one that must ramp upwith them. The path of acceptance on these principlesinvolves training, and trial-and-error, to ensureefficiencies are in place. People — Potential coding culture friction betweendevelopment and Quality Assurance (QA). Troubleshooting testing and determining why a testdidn’t work can involve time-consuming steps such asvetting the: CT