TiDB Hackathon 2020 TiDB Hackathon 2020 TiDB Hackathon 2020 TiDB Hackathon 2020 TiDB Hackathon 2020 This year is the fourth time that TiDB Hackathon has been held. The scale of participating teams is the largest ever, with a total of 45 teams from all over the world signing up, realizing global linkage for the first time. After 2 days of extreme challenge, many exciting projects emerged in the competition. In order to let more friends know the stories behind these participating teams, we will open the TiDB Hackathon 2020 Outstanding projects sharing series.
Ti-improve was a team formed one hour before the registration deadline. Four students, Zhou Qiang, Fang Zhuhe, Zhao Ylin and Li Dezhu, worked remotely in Beijing, Shanghai and Hangzhou respectively in this TiDB Hackathon and overcame difficulties within 24 hours with the help of the full set of AWS facilities. The basic framework of one-stop distributed database testing can greatly reduce the testing cost and expand the testing diversity. Finally, this project won the Most Potential market award of Huachuang Capital.
Xie Jia, head of enterprise software investment of China Innovation Capital, highly appreciates the Ti-Improve team. He believes that the project of Ti-Improve team not only improves the overall engineering efficiency in terms of technical implementation, but also has strong commercial potential in TiDB ecosystem. We take this opportunity to invite Xie Jia to interview with four students from the Ti-Improve team, and explore more exciting journey of TiDB Hackathon 2020 through the collision of investors and programmers’ thinking.
Q: What is the team process like at Ti-Improve? What’s the story of how the team got its name?
**Ti-Improve: ** Our team was formed temporarily. We came up with the idea of the project and found three other students to form a team one hour before the registration deadline. Due to the lack of time, we used the team name of last year, namely ti-Improve.
In the team, Zhou Qiang is mainly responsible for test infrastructure, Fang Zhuhe is the core developer of TiFlash, Zhao Yilin is the core staff of TiCDC, and Li Dezhu is the developer of HTAP. One hour before the deadline for the registration of this Hackathon, we formed a team under various circumstances and finally completed the project that we thought exceeded our expectations.
Q: Can you share the initial inspiration and opportunity for this project?
** TI-improve: **TiDB itself is designed to be cloud native, but also hope to have a better operation on the cloud, which is also the direction of future focus of development. Therefore, we wondered if hackathon could do a project related to a flexible test system that could easily import and export data on the cloud.
Through some simple research, we think this thing can be done, and finally we decide to fully combine enough components in AWS, such as lambda, Cloudformation and CloudWatch, plus pipeline, to become a particularly powerful platform. We can do all kinds of testing anytime, anywhere, and inject errors into things like Chaos Mesh, Fuzzing, and so on and put everything together organically.
We also have students who have been working on test infrastructure, and they will encounter various problems in the actual process of building test platforms. After investigating these components of AWS, we found that these components have been able to solve problems such as debugging, choreography and observability well, so we can directly see the effects behind them. One of the problems with TiDB is that while we do our best to make sure the test cases are large and broad enough, it’s actually very difficult to present the full picture of TiDB quality. In addition to development and testing, we can also do serverless TiDB, and even do some other things such as hooking up with the data lake.
In fact, after Hackathon, our team members also studied together what we could do with AWS infrastructure in the future. We also found another four or five new things that we could try and combine on TiDB later. Our initial thought was to see if it could all be done in one go on AWS and we didn’t have to work on it ourselves, which would make the development and testing iterations faster.
Q: There are 14 judges in this Hackathon. As an institutional judge, TI-Improve has always been a favorite. Could you share what you like most about this team and what impressed you most during the competition?
Xie Jia, ** HuACHuang Capital: ** This Hackathon is mainly an online collaboration. Ti-improve’s idea itself is very attractive to me. From my perspective, I hope to try to find some different perspectives from the other 13 technical judges. What I want to express is that innovation with smart business value is also a kind of hacking. Ti-improve may not be the most technically difficult project, but I quite agree with the business significance behind this idea.
What I think is very valuable is that it improves the lubrication of Customer Adoption. For example, ti-Improve is a tool for customers on AWS who want to use TiDB products and how they can get through usability testing as quickly as possible if it is possible to minimize the time and friction that customers spend on TiDB products. Secondly, the technical features of Ti-improve also well practice the concept of Cloud Native. The customer experience of Cloud Native should be able to fully coordinate the scheduling of various capabilities on the public Cloud.
Q: What technical difficulties did the team encounter during the competition? How is it solved?
**Ti-Improve: ** One of the issues that none of us expected to encounter during the Hackathon was that the biggest obstacle was the AWS permissions system, which was completely unexpected due to its complexity and large size. At the same time, we do not have the permission to operate our own account, and every step of operation will encounter two or three permission problems, which need to be solved one by one, which is very inefficient. Therefore, we went to the relevant person in charge to open the permission and met the person in charge to ask for leave. It was not until 11 o ‘clock on Saturday (the day of the activity) that the permission problem was completely solved. This was actually the biggest problem we had. It wasn’t a technical problem, but we wasted a lot of time on it.
Q: From the perspective of institutional judges, why did you choose this project? What are the technical and practical implications of this award?
Xie Jia of ** China Innovation Capital: ** We have more than 40 projects in total this time. In fact, as a judge of an investment institution, I was really torn when choosing a project, because I liked many projects very much, but I could only choose one project in the end, so I was really torn. Therefore, I divided my favorite projects into three categories of value, and then ranked them according to commercial value, technical difficulty, commercial urgency, proposal readiness and other comprehensive considerations:
I. Opportunities for independent productization and commercialization projects in TiDB ecosystem;
2. Projects that help IMPROVE TiDB’s underlying stability and other performance and have long-term significance for TiDB’s underlying core capabilities;
Iii. Projects that can accelerate and optimize TiDB commercialization adoption;
Ti-improve project completely uses AWS infrastructure to carry out TiDB testing, which, in my opinion, is typical in the third category. It not only solves specific technical problems, but also has great commercial value.
Since this year’s Hackathon is only 24 hours long, Ti-Improve only formed the team on Friday night. Do you feel any regret during this process? Where else could we improve if we had more time?
** TI-improve: ** I think there are several points that can be improved for us if we give enough time:
1) The system of the whole system can be built more perfect. Hackathon is a crab type, which opens a way for the architecture of the later test system. If there is more time to improve, I hope the whole process and details can be expanded and improved better.
2) I can do better in PPT and Demo presentation. If I can get through the road ahead quickly at the beginning, I can have more time to prepare PPT and Talk and do better;
3) should have more time to do research, even should be combined with many special things to the whole test system platform, as a matter of time before we actually on the same day are also reading a lot of documents, familiar with how to use, at the time of preliminary see a team is looking forward to for our last Demo, but due to time factor, finally only made it a basic process, I wish I had more time to make this thing a little bit more functional and imaginative, which is the biggest regret.
Q: Ti-Improve has won the most market potential award of China Innovation Capital this time. What is Ms. Xie jia’s expectation for the future of this project?
Xie Jia of ** Capital: ** just said regret, I feel a bit Rock Music, laugh ~
Ti-Improve expectations, I feel along this train of thought, really can do things or pretty much. Expect to see the ti-Improve project more detailed. On the other hand, I understand that the next important step for PingCAP is to have better user scenarios and better customer cases on mainstream public Cloud platforms such as AWS and Google Cloud. Therefore, I also expect ti-Improve to follow its ideas to perfect more similar functions into a system, which can be actually used by customers.
Q: Does TI-Improve have any future plans or expectations for the project?
** Ti-improve: ** Some basic testing facilities did not reach our expectations due to human resources problems before. Therefore, for this Hackathon, we hope to make up for these shortcomings by using AWS, so as to save labor costs and do more things well. For testing and testing systems, we think testing is very important, as a very core based software, stability is absolutely the most important.
We are looking forward to seeing this thing really run in the company in the future. For us internally, it not only greatly improves efficiency, but also provides us with another new idea for testing on the cloud. We are looking forward to doing tests on the cloud in the future.
Q: As an investor who majored in software engineering and is very experienced in investing in the field of enterprise services, what does Teacher Xie Jia think of Hackathon?
Xie Jia of ** China Innovation Capital: ** I have participated in some non-technical activities before, but they were basically to come up with a business model or idea and give a speech at the end, rather than a program that can finally be Run. So when I saw the actual effect of the Demo of the contestants in the final, I felt it was very infectious and impacting. Hackathon itself is a kind of improvisational creation process. This kind of innovation happens from the bottom up. I like this kind of spontaneous innovation very much.
Q: Besides your own project (Ti-Improve), which project did you enjoy the most?
** My favorite project is VR. It’s called Slippery egg. It makes everyone look so cool.
** My favorite project in my memory is to Improve the efficiency of Bug repair. The team name is Zhangyushao. I think it is very meaningful in terms of product stability and quality improvement.
** I like the first place project ‘or 0=0 or’ most, I think it is very imaginative, take a seemingly traditional thing here, and solve the problem well, I think this idea is really great, I can learn from them.
** I also prefer the first ‘or 0=0 or’, he attracted me is that he uses WASM to do this thing, WASM related stack actually I have always been concerned, but I also did not expect to combine TiDB implementation of UDF. I think it’s a good way to use this idea in the future.
Xie Jia, ** Huachuang Capital: In addition to Ti-Improve, I actually like a lot of projects. I originally listed 12 projects. I like the Ti-Graph team as well, because some types of data analysis are better suited to Graph databases.
About Huachang Capital
Founded in 2006, Huachuang Capital focuses on early stage investments in enterprise software, frontier technology, life sciences, and consumer upgrading, and uses its deep resources and experience to help the invested enterprises in strategic planning, business development and team building. Huachuang capital currently manages rmb8bn in renminbi and US dollar funds. Representative investment projects include: What is worth buying (GEM :300785), Daily Youxian, Wish (NASDAQ: Wish), Tongdun Technology, Tiger Securities (NASDAQ:TIGR), Ezbuy (NASDAQ:LITB), Another Red (MeituanhKEX :03690 Acquisition), Caiyun Technology (bytedance acquisition), Xiaoman Technology (Alibaba NYSE:BABA acquisition), PingCAP, Shenjian Technology (Selingsi NASDAQ:XLNX acquisition), Airlango (Meituan HKEX:03690 acquisition), Nreal, AI Pen intelligence, Blue Arrow Aerospace, Huake Precision, New Geyuan, Fubei Biological, bi Chen biological, Ying Shi biological, etc.