Shulou(Shulou.com)06/03 Report--

The data center is very popular this year, a little bit all of a sudden, and many users are lost by it.

This wave of Taiwan craze brought up by Internet giants seems accidental, but in fact it is inevitable. It makes us really realize the great value brought by the capitalization of data, as well as the importance of building data asset system and organizational structure adjustment in the future competition between enterprises and institutions.

When the data center has become the trend of the times, for the major traditional industries, we can not help but ask: how to create a data center suitable for their own business? It is true that the data strategy of Internet companies has its merits, but learning to walk in Handan may lead to losing the whole game. In fact, the data center is only a pronoun after all, to form a data asset management system suitable for its own business, and to provide "nourishment" and support for the foreground business through data capitalization. It is the core essence of traditional enterprises to build a data center.

Cold thinking: the loss and Progress of the data Center

As the saying goes, no matter whether it is an Internet company or a traditional enterprise, no matter what methods are adopted, there is no way to go the same way in the construction of the data center, but only by different paths.

Why do you need a data center?

The digital native enterprises represented by Internet companies are born digital, but the vast majority of Internet enterprises begin to focus on the construction of data center only after they have developed to a certain stage. Jia Xishun, head of the big data operating system, said bluntly that the typical characteristic of Internet enterprises is barbaric growth. When disorder expands and develops rapidly to a certain stage, it is often necessary to drive and control the business more efficiently through the data center.

And most of the traditional industry users, most of them have experienced years of information construction, from the past data warehouse, data management, to the later big data platform, data governance, is the process of steady evolution of enterprise data capitalization. At present, the traditional industry is in the process of transition to digital enterprises, and the data center is the next stage of data capitalization evolution, which is very important in the digital transformation.

As the saying goes, there is data before data. Traditional enterprises need the data center to drive the business and realize the rapid innovation of digitization. From a percentage point of view, the data center is an overall platform that integrates data collection, fusion, governance, organization and management, and intelligent analysis, providing data to the business foreground in the form of services, in order to improve business operation efficiency and continuously promote business innovation.

"if you compare the foreground to the application program of the computer and the background to the storage of the computer, then the middle station is the memory of the computer. The role of the data center station is to combine the fast response of the foreground with the contradiction of slow change cycle in the background, to 'memory' the stable general business capabilities in the foreground to the middle station, and to extract the frequently used capabilities in the background to the middle station. Now, the capacity of the 'memory' of the data center is becoming larger and larger, "Jia Xishun said of the data center.

Percentage point is one of the few domestic companies with rich experience in data center construction in many industries.

Take the percentage of users in a large group as an example, its business spans dozens of industries and has more than 2000 subsidiaries, and its business volume and complexity are huge. Almost every company has multiple business systems and data standards. In the process of digital transformation of the group as a whole, the data is aggregated, pulled, flowed, shared, and data is used to drive business applications and innovation. It has become the biggest challenge faced by the group in the process of digital transformation. "users like this are very representative in traditional industries and have a wealth of internal data, but they have not built a data asset management system and are unable to capitalize the data," said Jia Xishun. It is also impossible to support the business innovation of the group and its subsidiaries through data. "

In Jia Xishun's view, the data center will be a standardized support for enterprises at the data level, through the data center to support the front-end data applications that let a hundred flowers blossom, and a large amount of data generated by the application will flow back to the data center again to form a complete data closed loop and drive the formation of enterprise data ecology.

Okay, what's the data center like?

There is no doubt that there is no standardized definition of the construction of data center at present, and everyone's understanding is varied. At present, in addition to Internet enterprises, many enterprises are generally in the conceptual planning or just the initial stage of data center construction. Jia Xishun used three points to summarize the main challenges faced by enterprises in the construction of data in Taiwan.

First of all, most enterprises do not understand that the construction of data center is a long-term process, which is very different from the traditional IT project cycle. Business departments must promote and cooperate with the construction of data center from top to bottom. "in many cases, the construction of data center is not a technical issue, but a question of whether the concept is changed or not."

Secondly, enterprises need to ensure the construction of the data center at the organizational level, and need to set up separate departments or organizations. "even Internet giants such as Ali, Tencent and JD.com have turned the data center into independent departments to co-ordinate all the data of the group."

Finally, qualified teams, technologies and products are needed to complete the process of realizing the data center. "in fact, many traditional enterprises are relatively poor in terms of talent and technology reserves, which can be made up for by manufacturers such as percentage points."

So, what will a good data center look like? This is probably a topic that all enterprises and organizations will need to think deeply about next.

From a percentage point of view, the data center must be considered in the light of the actual situation of its own business. "it is also drinking water. If you are a family, you can dig a well to meet your drinking needs; if you are from a village, you can dig a large well and pump water; and if it is a city, you need to build a waterworks to meet the drinking needs of thousands of households."

Percentage data Zhongtai panoramic picture

However, regardless of the size and complexity of the construction process of the data center, the construction of the data center needs to focus on the connection of data, including data access, cleaning, processing, integration, and ultimately supporting the services of the application layer. " Regardless of the size of the data center, its link integrity is very necessary, and it is essential at the level of data standards and data governance. Only by unifying the data standards can we support the whole data ecology. " In addition, taking into account the growth of enterprises and institutions, percentage points also emphasize the scalability and flexibility of the data center, "the data center must be robust enough to support the flexibility of the business."

From a macro point of view, a good data center can invigorate enterprise data assets, establish a data asset management system that will be installed, form a data ecological closed loop, and effectively support the rapid innovation of data-based business. and help to break the internal inherent interests and departmental walls, promote the enterprise in the organizational structure level of reform, adapt to the digital transformation.

How to choose the data center service provider

The craze of data center platform has also brought fire to all kinds of data center companies, from companies that do big data marketing at the top to companies that do data warehouse and big data platform at the end, as well as traditional information software companies such as Yuyou and Kingdee. Even cloud manufacturers such as Aliyun extend from their own areas of expertise to the data middleware market, hoping to get a share of the data middleware market.

However, in the face of so many data service providers, how should users choose? The author believes that it will be more reasonable to consider it from three levels.

First of all, we need to pay attention to the accumulation of service providers at the technical level, such as data warehouse, big data platform, etc., to see whether they have corresponding technical experience and strong technical team.

Secondly, we need to pay attention to whether the data center service providers have sound products and tools, including products and tools of the whole system, such as data access, cleaning, ETL processing, visualization and so on. Whether these products and tools have been precipitated and tested by a large number of projects.

Third, we need to pay attention to whether the data center service providers have rich industry experience and business precipitation, such as government agencies, finance, retail and other industries are very suitable and need data center construction, although the data center needs of each enterprise or institution are very different, but the experience accumulated by suppliers in the same industry will help them build the data center of other users.

Finally, we need to pay attention to the delivery capability, a data center project, in which the data governance part is not very easy, involving data standard formulation, data pull, and so on, basically accounting for most of the dirty work. There is likely to be a situation in which some cloud vendors are unwilling to do it and small startups do not hold.

What is the future trend of data center?

"if your data scientists and data analysts still need to spend a lot of energy and time on data collation and processing after a data center has been built, then such a data center certainly does not represent the future development trend." Jia Xishun said so.

In a percentage point of view, the data center is born for AI to some extent, especially the current AI applications are gradually entering various industries, which need a large number of basic data sets to train and support. The establishment of the data center is conducive to the training of AI models and the innovation of AI applications. " The data center should be very easy to use, not just for professionals, but for a wider range of people to directly use tools and data to support product innovation. "

Secondly, AI technology to optimize and transform the data center will also become an important trend. "AI can make the data center more agile and efficient. For example, percentage has begun to integrate some AI technology into the underlying tools and products of the data center, which can quickly and easily process data in data tagging and data governance. In addition, scripts can be generated automatically through AI." Jia Xishun believes that the proportion of AI in the data center will become larger and larger. "in addition to the bottom, do some AI-level support for the upper-level business, such as NLP, question and answer robot, automatic translation, and so on."

Third, the customization of the data center will be easier. One of the big doubts about the data center is that the customization caused by the needs of users is serious, and the data center can not form universal products, which makes it very difficult to deliver the data center project. Jia Xishun said: "the products and tools of data Center will be more flexible and develop towards the direction of componentization, and can be built to meet the needs of users through building blocks. Personalized needs can also be quickly customized through components."

Overall, the data center will develop in a direction that is smarter at the bottom and easier to use at the upper level.

Write at the end

In the long run, the data center is actually not directly related to the size of the enterprise, and different sizes and types of enterprises need the data center. The data center will gradually become a basic capability, which will be necessary for enterprises in the digital age.

In any case, the construction of the data center will be a long-term, spiral process, and it will also have a far-reaching impact on the organizational structure of the enterprise.

As Frederick Brooks, the father of the mainframe, mentioned in the Man-month Myth: there is no silver bullet in this world. There is no way to build a data center once and for all. It is probably not realistic for enterprises and institutions to achieve results in a short period of time. Enterprises need to think more from a strategic level, and then cooperate with the selection of appropriate data center service providers in order to take a steady step on the road of data center construction.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.