Chapter 957 General Database

Zhang Yiming said that it is not unreasonable to poach people from Baidu.

Because there has indeed been some movement recently.

The "Pyramid" project that Baidu has been working on for a year seems to be running aground.

This starts with the history of technological development...

When the Internet was not developed and the amount of data was small, everyone used databases for data analysis. When the amount of data increases to a certain extent, the traditional method will not work, and it must be applied to distributed technology.

Baidu is considered one of the earliest companies in China to set foot in this field, similar to Ziweixing.

The project team was launched in 2006, using distributed technology to increase search efficiency and ensure the fastest and fastest search response.

Later, the amount of data continued to increase.

The concept of big data is emerging.

That is, last year, Baidu prepared to build a technical research group to develop a domestically developed big data system, namely the Pyramid Project. This time period is similar to that of Silicon Valley giants such as Microsoft, Amazon, Yahoo, and IBM, and they are launched almost at the same time.

But the Internet is changing too fast.

Only one year later, the situation has changed.

The big data open source product Hadoop developed by Yahoo has become the mainstream of the world's technology companies, and the Hadoop big data system has been quickly established by using the open source mechanism.

Both IBM and Yahoo have given up self-research and devoted themselves to Hadoop, including Adobe, America Online, MySpace and other companies.

What a great open source product!

Developers all over the world work together and improve together. Share with each other and grow together.

Not to mention saving time and effort, it can also save a lot of research and development funds.

At this time, differences emerged within Baidu.

Do you want to join Hadoop like those Silicon Valley giants? Looking at the current trend, maybe in the next few years, giants such as Amazon, Microsoft, and Facebook will also move closer.

Relying on open source products is more cost-effective.

However, in the name of domestic independent research and development, do you still want it? Baidu's good reputation in the past few years is largely due to its autonomy in the field of technology.

On the left is the national strategy; on the right is the cost of the enterprise.

Is it left or right?

In the previous life, Baidu persisted for a while until it disbanded the team in 2009 and chose enterprise cost. Because the team leader at that time was full of idealism, fighting for reason and full of confidence, struggling in Baidu's internal struggle.

This person is Yang Zhenkun.

This life is different. Yang Zhenkun has long been under Ziweixing's command and led the team to develop a distributed database.

Without Yang Zhenkun's persistence, how long can Baidu last?

You know, it's an economic crisis right now.

Baidu is a listed company, and its stock price has fallen terribly, and it has been under tremendous pressure from investors. In terms of business expenses, of course, it can save a lot of money.

Li Yanhong is a scientist and also has a passion for science and technology, but there is a prerequisite, the enterprise must survive. The profitability of the enterprise is poor, the management is unhappy, and the investors are even more unhappy. He's not very good at maneuvering, if he doesn't handle it well, he may have to betray his relatives.

Even if Li Yanhong wants to do it, but the management does not support it, and the investors are not optimistic. He wants to force a decision, which requires great courage.

In the afternoon, Zhou Buqi participated in a technical committee and talked about it.

Everyone is ambitious.

Shen Xiangyang, Cheng Binghao, Wang Jian, Yang Zhenkun and others are looking forward to it.

"Pyramid" is an idealistic project.

The domestically developed big data system, if it is made, will definitely be an indomitable product. Therefore, even though this project is very hard and takes a lot of hard work, it has attracted many technical experts with sentiments, and they have to make some achievements for their ideals.

Once a project is outlawed, it's over.

It is tantamount to ruining the ideals of these technical elites and scientists.

The result can be imagined.

Can these people still stay in Baidu? Ideals are shattered, that is sad ah. They don't have to worry about finding a job, and they must end up leaving collectively in frustration.

Moreover, even if they wanted to stay, they couldn't stay under the power struggle.

One technical vice president wants to be a "pyramid", and the other technical vice president wants to switch to the Hadoop system. This is an internal power competition. Whoever wins will become the new chief technology officer, which is a bit of a life and death struggle.

Winner takes all, in order to ensure their status, what should we do?

Of course, taking advantage of the opportunity of banning the Pyramid project, all the competitors' direct troops were cut off!

You know, in order to carry out this pyramid project, Baidu has drawn out the most elite and core technical force to form such a project team.

Once the project is abolished, it will cause a huge loss of technical talents to Baidu.

Ziweixing can take advantage of the situation and take over the whole set.

Especially Yang Zhenkun, whose distributed database involves self-developed big data processing, needs similar talents the most. With the online operation of Feitian database, many shortcomings have also been exposed.

The single data processing capability is indeed very powerful, but there are defects in the collaboration, coordination, and scalability of the entire system. At present, it can only be deployed on less than 500 servers, and more technical accumulation is needed.

In this regard, Xu Liangjie helped to explain: "It took Oracle six years to make the database from research and development to maturity. Our Feitian database is only two years old, and it will take longer to improve and perfect it."

Shen Xiangyang said: "The distributed database is a future-oriented database, and it will definitely become the mainstream of the database market in the future."

This is very obvious.

The future will be mainstream, but now it is non-mainstream.

These days, it is still the world of relational databases.

This is also an important topic of today's technical committee, about the next important project to be carried out in the cloud computing business - cloud database.

Zhou Buqi thought that his own Feitian database had been made, and it would be fine if he changed it a little bit and turned it into a cloud database.

But reality doesn't work.

The birth time of Feitian database is too short. At present, it has only completed the design of some basic functions, and there are still too many details that have not been studied. Like scalability. Strong scalability should be a major advantage of distributed databases, but Feitian Database has not yet studied this aspect, and its scalability is not as good as Oracle. Oracle can expand to 5,000 servers, and Feitian has a cap of 500 servers. there's still a long way to go.

There is a more important point, that is, compatibility and diversity.

In this regard, Dr. Wang Jian gave a wonderful explanation——

"The Feitian database is like the six-meridian sword. It is the top martial art, but it may not be suitable for everyone. Duan Yu, a talented wizard, can practice, but those with poor qualifications can't do it. You can only learn Qishangquan , Fuhu Quan and other martial arts. Some people have high aptitude, and they may not like the Six Meridians Excalibur, Qiao Feng likes the Eighteen Palms of Subduing Dragons, Dongfang Bubai likes the Sunflower Collection, Zhang Wuji likes the Nine Suns Divine Art...no one likes it. Bad. That's diversity."

"The field of databases is diverse, and there are various databases of different styles, such as native multi-mode databases, time-series databases, graph databases, tabular databases, and various styles of relational data. Just like martial arts cheats, there are thousands of different styles in the world. people have different choices.”

"The Ziwei cloud database should be built as a Buddhist scripture pavilion to accommodate as many of these different types of databases as possible. Our self-developed Feitian database is just one of them. After the cloud computing platform is built, there will be success. Thousands of companies have settled in, and it is impossible for them all to use the Feitian database. We need to store enough martial arts cheats to meet the needs of different martial arts enthusiasts."

Zhou Buqi was very happy and felt that his reasoning was very clear.

Shen Xiangyang said: "Distributed databases can represent the future, but not now. Our cloud computing business will be launched within a few years, and at that time, we can't count on Feitian. The top priority is that Ziweixing still needs to develop a mature and stable database. An autonomous and controllable relational database, as the main database on the cloud platform."

Wang Jiandao: "Well, it must be based on MySQL." Then he added to Boss Zhou, "This is an open source database with many shortcomings. However, it has been open source for so many years, and developers all over the world have optimized it to the greatest extent."

Yang Zhenkun's face was unnatural, and he felt abandoned.

But there is no way.

The time for Feitian distributed database is too short, and there is still a long way to go for purely domestic self-research. Even if it is used internally, it can only be used in a certain field.

A very important point of the database is its versatility, which must be applicable to multiple scenarios. However, the Feitian database itself has no compatibility, and cannot use the syntax and models of many traditional databases...it is very inconvenient to use.

This is like making a mobile phone operating system compatible with Android.

Incompatible, no matter how strong the technology is, it can only be used in a small area, and it is difficult to promote it on a large scale. Android, which relies on open source, is more popular. Other companies like it, and so do their employees. Moreover, compatible with Android, you can use the dazzling application services in the Android ecosystem.

These application services have been developed by countless companies for many years.

If they are all self-developed and done by themselves, it is almost impossible.

In other words, the Feitian distributed database is very powerful and the technology is very advanced, but it can only serve a certain special field, such as the gold coin wallet system with extremely high throughput. Or if YikYak wants to use this system, the Feitian database must be developed specifically for YikYak and designed specifically for YikYak's needs.

Insufficient versatility.

Ziweixing needs to develop a general-purpose database on the cloud platform. It can rely on the open source MySQL to have more powerful and richer functions.

For any new business, you can directly use this database. If you use the Feitian database, it will be more troublesome. It needs to be specially designed and built for this business, and it will be maintained and upgraded separately, which will affect the efficiency too much.

Zhou Buqi heard the general idea in a daze.

The basic situation is understood.

If Ziweixing only does friends.com or micro-point search, then the Feitian database is feasible, one-to-one special service, tailor-made, this will be the best database.

But Ziweixing has too many businesses, covering all aspects.

Feitian's team is only two to three hundred people, it is impossible to handle so many businesses. In addition, the cloud computing platform is aimed at thousands of enterprises, and each enterprise has different database requirements.

The Feitian database is not enough.

Too young and too young, it takes longer to grow and develop.

Ziweixing needs a general-purpose database as the main database of its own business, and at the same time push it to the cloud, and promote it to other enterprises as the main database of cloud computing.

This general-purpose database can only be the open source MySQL.

Wang Jiandao: "At present, there are 278 people in the cloud computing business, less than 280 people. The main job is to design the cloud operating system and develop based on big data. If you want to go to the cloud database business, you need to invest another 250-300 People. Add another 60-70 million in expenses every year."

Zhou Buqi waved his hand, and said with a smile: "Ten billions have been invested, and there are still tens of millions left? Since there is a need, then do it. Old Shen, you can coordinate in the middle to see whether it is a transfer or a recruitment, and deal with it." .Cloud computing is the core strategy, there is nothing to hesitate, if you want to use it, go decisively!"

Dr. Wang Jian was very happy.

Yang Zhenkun sighed.

The general trend is still unstoppable.

Compared with the cloud computing platform, Feitian database is just a gadget.

Zhou Buqi said: "Of course, the Feitian database can't be stopped. This is also a key project and needs to be supported vigorously. Didn't you say that distributed databases are the future. You can't always think about overtaking on curves, you have to overcome difficulties , starting from scratch and doing research bit by bit.”

Yang Zhenkun looked a lot better, and asked, "Where is Baidu?"

Zhou Buqi pondered for a while, "This matter can't be rushed. The country is going to give awards to Baidu and Ziweixing. It is said that it is recently, but I don't know when it will be. It may be next month, or it may be the next month. They The efficiency... However, before the award is issued, I don’t think they will terminate the pyramid scheme, this is their face. After the award is issued, let’s take a look.”

Yang Zhenkun raised his eyebrows, "Is there really a chance?"

Zhou Buqi smiled and said: "The opportunity is great. When the financial crisis breaks out a little more seriously and Baidu's stock price drops a bit, they won't be able to handle it. Li Yanhong's character is a weakness. Even if he can cover the sky with one hand, he doesn't have the courage. He cannot withstand the pressure of investors and management. In the face of official policies, investors, management and personal reputation, he must make a trade-off to balance. Under the economic crisis, survival is the most important thing.”

Yang Zhenkun was very happy, "If the group of technical elites from the Pyramid Project can be taken over, the prospect of Feitian Database will greatly increase."

Zhou Buqi said decisively: "Opportunities and risks coexist. The trend of the financial crisis is already obvious, and it is a bit unstoppable. Other companies are avoiding risks, but we are looking for opportunities here. Not only Baidu, Microsoft, Google, Oracle, and IBM are all listed companies, and their operations are under great pressure. You can go to their research institutes to poach people. Domestic counterparts must exercise restraint, and go to foreign companies to poach people boldly.”