Series of articles:
Ok, the
What is the middle station? What problem is it solving?”
The next”
Misunderstanding about China and Taiwan and the selection of China and Taiwan interpretation”
The beginning of the story
This concept, which was first proposed by Alibaba in 2015, is an extension of its strategy of “big in the middle, small in the front”. Alibaba, Tencent, Baidu, JD.com, Meituan, Didi and other Internet giants began to adjust their organizational structure from last year to this year, with the intention of building the zhongtai……
And one sunny afternoon last week, I was eating a cake. The boss suddenly pulled all of us into the meeting room and said to us earnestly that we would make a data center.
Although a meeting down, even he did not say clearly what is “in the middle”? But if you don’t understand, don’t say anything. Until the moon cake is finished, let’s compare the situation without “middle platform”
Project status
There are a lot of internal/external/support, but no matter how complex the internal project is, the general structure is “user front” and “management back”.
Users at the front desk
User-oriented, directly interactive, pages focused on design/interaction,Generate data exchange with the server to guide the user through the business process. Such as:
Management background
The configuration management system for operation personnel, the background provides some simple configuration for the foreground.Such as:
The relationship between user foreground, management background and users is as follows:
Under the traditional mode, the project iteration cycle is basically monthly and quarterly. Long development cycles also mean that when requirements change, either 996 or delivery is delayed
And the projects are relatively independent, with many reinventing the same wheel. Let the project more and more bloated at the same time, also let the development efficiency is less and less.
But the reality is that the Internet has entered the second half and the competition among enterprises is becoming more and more fierce today. The product project can not be fast iteration, low cost of the consequences of trial and error, equivalent to put the enterprise at a certain competitive disadvantage.
In order to solve the above problems, the concept of “middle Taiwan” arises at the historic moment
Supercell, the pioneer of the middle Platform concept
Tolerate failure, even celebrate it
Supercell, failure is never a shameful record, but rather an incentive to progress. After a game launch fails. The response from its managers was “Great, the game failed and proved that we were eliminating a wrong path”. A unique “celebration of failure” is rooted in its culture, paananen says: “We built this company on learning from our failures. The faster we fail, the faster we learn and the better we get.”
Failure is the mother of success. Can you really learn from failure the information you need to learn how to succeed
Deep use of cloud computing/microservices and other technologies to upgrade the value of individual employees
Supercell, which uses Amazon’s AWS service, processes terabytes of data every day, and these logs will be used to improve the game experience. Sufficient and unified public business modules, to assist the team to avoid duplication of work, focus on gameplay, game experience creation can quickly build products. While increasing the value of individual employees, it also reduces the cost of trial and error
Small but small teams
. Small teams mean that the cost of trial and error is not too high, time and direction can be adjusted and changed in time to meet the needs of a fast-changing market, the aircraft can be more agile, and once the right target is found, they can focus on expanding their success
Information sharing
Summary should have an idea
Middle Taiwan to solve the problem
From a technical point of view, the problem that the middle platform solves is that there are multiple projects and they are relatively independent, resulting in the need to re-build wheels, such as: file upload/order/payment/search module… The research and development cycle caused by long, programmers 996 can not flexibly respond to business changes
At the business level, the middle office solves the problem of long r&d cycle caused by technical duplication due to relatively independent projects/always slow to meet market demand (inflexible)/high cost of trial and error/not good for innovation
The middle way to solve
- Form services through a unified common technology module. When the service is needed again, it is called through the interface to avoid the duplication of the wheel and the extension of the development cycle
- Define business processes and encapsulate them into common business process modules. Reuse directly the next time you go through the same business process. Reduce the cost of trial and error, conducive to innovation
- In that case, the technology will not be developed as a project, but as a service formed by these “common modules”, forming the middle platform
With all that said, here’s an example:
China instance
Their games look very different, but they have a lot in common. In business, the common things are payment systems, user systems, etc. In technology, the common things are game engines, internal development tools, etc. And these common resources can be provided by a powerful “middle station” :
The architecture of the middle office changed not only the structure of the project, but also the organization of the R&D team. SuperCell calls this efficient form of organization “clans.”
Following closely behind, domestic Internet companies have also started their own china-Taiwan strategy. Alibaba has put forward the strategy of “big middle platform, small front desk” :
In the picture, the common business of alibaba’s many product lines has been sunk to form various business centers in the Middle Platform, while Aliware is Alibaba’s technology middleware platform to provide technical support for the major business lines.
The above is the specific meaning of the concept of middle Taiwan.
So much water finally water to personal feelings
The concept of “middle platform”, in a small way, is “micro-service”. But the middle desk needs to be treated in a product management way. Because the external services provided by the middle office need to be constantly iterated to adapt to the needs of the business, rather than waiting for the business to raise the demand. Yeah, technicians need to know business, too.
What’s the difference between product and project management? This you have to find products to exchange ideas, play next martial arts ~
Ok, the
What is the middle station? What problem is it solving?And that’s the end of it. Welcome to read more
The next”
Misunderstanding about China and Taiwan and the selection of China and Taiwan interpretation”Is aimed at why need in Taiwan, what kind of Taiwan in Taiwan is a good article, worthy of reference
Read the reference
Enterprises have followed The example of Ali built in Taiwan, after all is blindly follow the trend or do not die? |
Mp.weixin.qq.com/s?__biz=MjM… |
I take a fancy to |
Mp.weixin.qq.com/s/fQ98fe3XH… |
Manga: What is chung Tai |
Mp.weixin.qq.com/s/rF7_xJBq4… |
Nuggets: Data center |
juejin.im/search? Quer… |
Read through “Alibaba Data in Taiwan practice”, in the end what is brilliant? |
Juejin. Im/post / 684490… |
Zhihu: How can zhongtai respond quickly |
www.zhihu.com/search?type… |
Business in Taiwan exploration and practice: the fundamental problem of software |
zhuanlan.zhihu.com/p/59867439 |
What is the middle station? What is not the middle station? All middle stations are service centers |
zhuanlan.zhihu.com/p/77097815 |
How to build middle Taiwan? Organization, supporting technology and methodology of zhongtai construction |
zhuanlan.zhihu.com/p/77362869 |
What is human resources in Taiwan mode? |
www.zhihu.com/question/33… |
China: appropriate letter agile | share transcript data middle construction practice |
|
The cornerstone of OPPO Data Center: Building real data warehouse based on Flink SQL |
|
Does your business really need a “data center”? |