Time to say goodbye to Jenkins

01. Technology is a product of the times

I am Lao Guo, the founder of KodeRover , and you can call me Grant . Since returning to China to start a business in 2018, my team and I have built Zadig from scratch into the favorite cloud-native CI/CD continuous delivery tool for domestic developers and the most trusted cloud-native DevOps platform for enterprises, attracting thousands of companies. Moving from Jenkins to the Zadig platform. I have always firmly believed that technology is a product of the times, and I have witnessed this with my own eyes from the more than 20 years I have worked in Silicon Valley in the United States. In 2001, after I graduated from Stanford University, by chance, I joined Google, which was developing rapidly in the ruins of the economy and had only more than 100 people. This experience made me obsessed with software technology entrepreneurship. In 2006 I joined a search advertising company founded by earlier Google engineers and was acquired for $60 million just 9 months later. After that, he participated in the e-commerce start-up of TRX, a global super fitness brand, and witnessed the rise of Amazon e-commerce and the "cloud service" market they created. Immediately afterwards, four of our old Stanford classmates rented an office on Palo Alto University Road, the center of the universe in Silicon Valley, and started a video transmission technology business. At one point, we achieved 25 million MAU, and we were neighbors with Facebook, Tesla, and Paypal within 500 meters. In those years we have witnessed the transfer of applications to the mobile terminal and the cloud.

With the transfer of the world's economic center to China, I decided to return to China in 2018 and work with my partner and software engineering expert Li Qian to build a respected basic software company from China. In this article, I will use our product Zadig and the traditional domain tool Jenkins as examples to share with you my thoughts on technological innovation and the development of the times, hoping to bring some enlightenment.

02. Losing developers means losing everything: Jenkins inventor leaves Jenkins project

He decided to develop a tool that everyone could easily build: this became Jenkins. Jenkins clearly guessed the developer's mind correctly, and it has become the most popular build tool in the world for the past 20 years. However, with the rise of cloud native, it began to seem powerless, and it became increasingly difficult to guess the minds of today's developers, and Jenkins' experience for developers is getting worse and worse! In 2018, KK admitted that Jenkins has a bunch of problems, such as service instability, fragile configuration system caused by thousands of plug-ins, inability to plug and play, slow development speed, inefficient test quality, and inability to perform engineering analysis. process data, etc. Ultimately, KK left the Jenkins project and his company in 2020.

03. A girl who loves code: Jenkins experts give up using Jenkins

In 2017, Li Qian, a 27-year-old software engineer, was in charge of the engineering team at Qiniu Cloud, including the development and maintenance of Jenkins ecological tools. In the same year, Li Qian was invited as a Jenkins expert to share her experience at the first Jenkins MeetUp in China. But the fact is that in the face of thousands of R&D collaborations and high-frequency delivery scenarios, she realized that there are huge limitations in using Jenkins to manage software delivery: defects such as low efficiency, low quality, frequent accidents, and difficult maintenance are in the nick of cloud computing It becomes more and more prominent in the service scene.

From 2015 to 2018, with the expansion of Qiniu Cloud's multi-products and the exponential increase in delivery pressure, the engineering team led by Li Qian found that the Jenkins system could not bear the heavy load, the operation efficiency was low, and the maintenance cost was extremely high

At the same time, Kubernetes has gradually emerged as the de facto standard for the next-generation operating system, showing a powerful "cloud-native" capability of declarative resource scheduling, and bringing feasibility to the quality engineering of the whole process of software delivery. It was at the critical age of 27 that Li Qian started to develop a brand new delivery tool based on containerization, and promoted and tried it within Qiniu Cloud. Zadig today.

04. Jenkins is right, it's just old

Zadig and Jenkins were solving completely different problems from the start.

Jenkins was born 20 years ago to solve the problem of build efficiency in collaborative development. It works by managing software build compilation and relying on plugins for distribution. However, as KK pointed out, Jenkins has problems such as unstable service, a large number of plug-ins leading to a fragile configuration system, difficulty in getting started, slow development speed, and inefficient test quality. Although Jenkins seems to be running normally on the surface, it actually puts a huge burden on the R&D team, far behind the level of cloud-native pioneers in the same industry.

Jenkins is right, it's just old and no longer adapts to the evolution of collaborative delivery requirements of the times!

In contrast, the young Zadig cloud-native DevOps platform has risen rapidly, focusing on improving the efficiency and quality of the entire software delivery process. Zadig has demonstrated clear advantages in terms of environmental management, ease of use, quality engineering, out-of-the-box use, and development speed. More and more engineers are beginning to learn and replace the Jenkins system with Zadig to create a digital R&D highway for enterprises. On this new "highway", instead of "building cars" and "building roads" at the same time, developers can focus on development work. Derrick, CEO of the new retail leader Feima, once said with emotion: "Zadig helps us liberate productivity, and the efficiency may not be increased by several times, but by more than ten times. It brings not only results, but also the improvement of personnel capabilities. , and this is the most important."

05. Zadig: Surpassing Jenkins and following the times

In the past two years since it was launched independently and open-sourced, Zadig has grown from nothing to become the first choice of thousands of domestic companies and has won the favor of a large number of developers. I have summarized the main reasons why everyone loves Zadig:

Emphasize developer experience and improve team capabilities: Zadig integrates team capabilities into the platform to achieve high human efficiency, high quality, high efficiency, and low cost. People can interact efficiently on the Zadig platform, sink the complexity of collaboration to the platform, and improve the human efficiency of the entire organization. This breaks down organizational, process, and system silos, creating a growing organization.

In contrast: The fragmented nature of Jenkins leads to the prevalence of empiricism. People manually switch between countless platform tools, which is time-consuming and labor-intensive, with low delivery quality, low efficiency, and high delivery costs.
Enhanced integrated engineering capabilities: Zadig covers efficient construction, environmental management, and high-quality releases throughout the entire R&D process. Supports multi-service parallel deployment and release, cloud-native build environment and operating environment, as well as infrastructure docking and enterprise-level SSO/rights management and other functions.

In contrast: Jenkins has a low level of engineering, security risks, low operating efficiency, and high management and maintenance costs.
Customized solutions on demand: For example, in terms of software supply chain security, Zadig combines static and dynamic code scanning, software component analysis (SCA) to generate SBOM lists, and effectively builds DevSecOps solutions. For quality engineering systems, Zadig integrates existing testing tools to provide unit testing, integration testing, system testing, performance testing, and layered automation to build continuous testing solutions.

In comparison: Jenkins and other tooling products can barely do this.
Strong openness: In addition to the support of built-in mainstream tool chains, it also provides service-oriented access methods such as using OpenAPI and custom development tasks, which can be customized by enterprises according to their own processes, thereby reducing the need for enterprises to manually build and use fragmented tool chains. cost.

In contrast: Jenkins and other tool products only solve specific problems in a single dimension, and fragmentation after expansion is inevitable.

06. As a technical leader, you need to settle accounts

Today, advanced companies have gradually given up on the Jenkins-based self-developed DevOps platform. There are two reasons behind this phenomenon: on the one hand, cloud-native technologies like Zadig are mature and reliable; on the other hand, the cost of building a platform by ourselves is out of control. Under the combined effect of these two factors, why invest millions every year to build a wheel that may not necessarily run?

Of course, there are still many technology managers who don't realize this.

Some people in charge are still building wheels blindly: they think that "Jenkins can realize all the functions of Zadig. If we implement it ourselves and do self-research, the cost will be low and the intellectual property rights will be independent."

My answer is: the continuous plug-in and secondary development based on the Jenkins system has aggravated the fragmentation of enterprise R&D management, and also wasted a lot of human resources on redeveloping useless wheels. However, today when cloud native and artificial intelligence are about to explode in various industries, the "intellectual property rights" and "technical capabilities" accumulated by the Jenkins system have instead become obstacles to the improvement of the capabilities of development, operation and maintenance, and testing teams, hindering the development of enterprise software products. The speed of iteration and the ability to adapt to the market will lose its attractiveness to outstanding technical talents and valuable market opportunities. I recommend at least getting the team to start using the open source Zadig, otherwise it's really out of date.

Some people in charge are still calculating a messy account: "If Zadig is used, then our team "just" does the configuration work on Zadig, which does not reflect the team's ability, and the labor cost and commercial purchase cost are also incurred. The final effect is similar to using Jenkins directly."

My answer is: "If a worker wants to do a good job, he must first sharpen his tools." This is like saying that high-speed rail and green trains can travel from Beijing to Guangzhou. Zadig has provided a platform like a high-speed rail. The key lies in how to use the advantages of this platform and empower it to the R&D team to maximize the team's productivity. This is not just "doing configuration". At a price of 100,000 yuan per year, you can let Zadig support a 100-person R&D team to meet various development needs, increase R&D productivity by 1 to 3 times, reduce construction computing resource consumption by 60%, and reduce O&M manual operations by 80% , so as to liberate the operation and maintenance team, allowing them to focus on more core tasks of the enterprise, such as higher-level innovations such as online stability and release efficiency. These are ZadigX's quantifiable return on investment (ROI). It is your responsibility to focus on these opportunities.

Some people in charge are still waiting and wandering: "Our team is not familiar with Zadig's source code, and we can't solve it in time if something goes wrong. Zadig's operation is not black-box enough, so I might as well use Jenkins, which I'm already familiar with."

My answer is: Zadig is 100% open source, and the source code is publicly available. The so-called "Zadig's operation is not black-box enough, and I'm not familiar with Zadig's source code", we can ask those colleagues who are familiar with Jenkins, how much do they know about Jenkins' source code? It is more because of the formation of habits and the incompatibility of "switching cars from horse-drawn carriages". At the same time, part of the ZadigX enterprise version is not open source, but the source code can be provided to strategic cooperative enterprises through special agreements. In fact, source code is never a problem. Facing the future bravely, solving more practical problems, and creating corporate value are issues that every team member should think deeply about. I highly recommend letting team members get hands-on with a project once and for all, they will find the experience incredibly smooth and convenient.

Let's end with a few fun facts

Japanese engineer Kosuke Kawaguchi (KK) was working for Sun Microsystems when he wrote Jenkins, when the project name was Hudson. In 2009, as Sun was acquired by Oracle, the ownership of Hudson was disputed. Finally, in 2011, the open source community chose to part ways with Oracle's Hudson and independently established the open source project Jenkins, which has continued to develop to this day. Sun is one of the most influential technology companies in Silicon Valley in the 80s/90s, they invented the world's first Sun Unix Workstation and the Java language. With the $7.6 billion acquisition by Oracle in 2009, Sun's headquarters in Silicon Valley has also become Facebook's global headquarters today. Sun founder Andy Bechtolsheim was also Google's first angel investor, investing $100,000 in Google in 1998. As for the video transmission technology company I participated in, its core developer and founder, Bryan Yang, joined Kealia, a company co-founded by Andy Bechtolsheim and Stanford University professor David Cheriton, after graduating from the Department of Computer Science at Stanford University, and was later acquired by Sun. The exit from Kealia also became Bryan's first pot of gold in video transmission technology entrepreneurship!

The main developer of Zadig is Li Qian, a former Qiniuyun software engineer. From 2015 to 2018, with the expansion of Qiniu Cloud's multi-products and the exponential increase in delivery pressure, the engineering team led by Li Qian found that the Jenkins system could not bear the heavy load, with low operating efficiency and high maintenance costs. With the continuous maturity of Kubernetes, Zadig was initially used as an environment management tool. After five years of development, Zadig gradually evolved into a cloud-native software delivery platform with a complete quality engineering system by shifting quality verification to the left. What's interesting is that many cloud vendors that are considered competitors have already adopted Zadig services for their internal teams. Byte's Feishu, Tencent, and Huawei are already Zadig's commercial customers before Zadig's open source in 2021. And Zadig never defines who he is. As Lotus Operations Director said: "In the market, Zadig has no real competitors. In terms of technical decisions, it has never embarrassed me. Its value lies in improving efficiency, and the user reputation is very good." Today The engineers affectionately called Li Qian "Sister Qian", and said that they were running naked without the protection of Sister Qian when writing code. And Sister Qian herself is a post-90s generation.

broadcast around

ZadigX Custom Sticker & T-Shirt Kit

How to get it: Share your experience of switching from Jenkins to Zadig in the comment area, and draw three award-winning students to win

Read the original text: https://mp.weixin.qq.com/s/Pedr4nXxG4HR6DHhJRkIlA