The concept of data center was first invented by Ali. At the beginning, Ali proposed the establishment of data center with a very clear purpose, which was to solve the data island between various departments. Data standards and caliber could not be unified, resulting in a great waste of data resources.
After more than ten years of development, the heat of the data center has gradually cooled down from the fire to the bottom. How will the data center develop in the future? Today, Fan Soft jun shares an article with you, the content is from the public account “Talk data” :
What is data center?
According to the problems to be solved in data center, we can determine the ultimate goal of data center construction. Data center is a kind of IT system, and the ultimate goal of IT system construction is to serve enterprises, so the construction of data center follows the business-oriented path.
Although enterprise development goals are diverse, such as alibaba’s goal is to “so” there is no such thing as a difficult business, tencent’s goal is “to the technology rich Internet user’s life”, but the big goal have a common goal, that is most effective to achieve the rational allocation of resources and use, to create the largest corporate profits, simple is fine operation, Open source and reduce expenditure.
From the earliest accounting system, to the information construction in the age of computer popularization, to the present big data, digital transformation and intelligence, all serve this goal. Especially in the network era, many industries form a winner-takes-all situation, enterprises need to be one step ahead of their competitors, take the initiative in the fierce market competition, and obtain higher profits.
Therefore, the ultimate goal of ** construction data center is to achieve “rapid market response, refined operation, increase source and reduce expenditure” through efficient digital operation. ** Digital operation is a necessary means for enterprises to gain relative advantages in market competition.
The relationship between data center and traditional big data platform
So, what is the difference and connection between data center and big data platform? See the picture below:
The figure above shows the relationship among information system, data warehouse, traditional big data platform and data center, where the arrow indicates the main flow of data. We can understand that traditional big data platforms and data warehouses are the data sources of the data center, and the construction of the data center is to better serve business departments.
What are traditional big data platforms?
Core functions of traditional big data platforms include:
- Basic big data capability layer: Hadoop, Spark, Hive, HBase, Flume, Sqoop, Kafka, Elasticsearch, etc.
- ETL pipeline built on big data components, including data analysis, machine learning programs.
- Data governance system.
- Data warehouse system.
- Data visualization system.
But data center should be a superset of big data platform. In our opinion, on the basis of big data platform, data center should also provide the following system functions.
1. Global data application asset management
Data application asset management here includes data and applications across the entire ecosystem.
2. Global data governance mechanism
The data governance mechanisms provided by the Data Center must allow the business units to iterate autonomously, but only if there are globally consistent standards.
3. Self-service and multi-tenant data application development and release
The data center emphasizes empowering business units, and business people need a self-service development platform that can adapt to different levels and capabilities.
4. Data application operation and maintenance
Users should be able to easily self-publish their data applications to production systems without having to go through a dedicated data team.
5. Data application integration
New data applications should be integrated at any time.
Data as a service, model as a service
The results of data analysis, whether from statistical analysis or machine learning-generated models, should be quickly distributed without code and available for use across the organization.
7. Data capability sharing and management
Most of the data capability should have a complete sharing management mechanism, convenient and secure sharing mechanism and flexible feedback mechanism. It’s the individual who ultimately decides how the data is used, and they need a mechanism to access that information, so you have to have that sharing mechanism within the organization in order for the data to actually be used.
8. Perfect operation indicators
Data center emphasizes measurable data value. Therefore, it is necessary to have certain operational indicators for the way data is used in the system, the frequency of data use and the final effect, so as to verify the value of data and the efficiency of data center project.
Data center after all is “silver bullet”?
In recent years, the concept of data center is very hot, many manufacturers have promoted their own data center related products or solutions, and the industry is also very optimistic about the prospect of data center. So does this mean that the data center is the “silver bullet” that will solve everything?
There is no silver bullet to cure all ills.
The answer, of course, is no.
The success of data center is built on the basis of information, without a perfect information foundation, enterprises can not fully understand enterprise business, more difficult to obtain useful information. In addition, Data Center provides rapid insights into existing products and markets, and improves existing products and operations. In other words, data Center can help market development, develop new business models, and accelerate the speed of iteration. However, the ultimate realization still depends on the creativity of the data Center team.
Generally speaking, enterprises with multiple business divisions and multiple product lines need to form data sharing and reuse in multiple product lines to maximize the input-output of data center. When multiple product lines and business units form data synergy, the role of data will be maximized.
It needs special attention that the construction of data center is not the ultimate goal of enterprises, enterprises should not build data center in order to build data center, not to blindly follow the trend. The ultimate goal of ** Data Center is to help enterprises realize digital operation and become data-driven enterprises.
What are the core capabilities of data Center?
The core idea of building a data center is to empower business units and provide better data capability tools so that business units can quickly gain business insights through the functions provided by the data center to quickly provide data-driven business products. Therefore, without business application, the construction of data center is a castle in the air.
When we plan for the construction of the data center, we should have business application scenarios, and subsequent iterations must be driven by real business requirements. It is worth noting that although we emphasize business drive, the overall planning and global data specifications provided by the data center are essential. Otherwise, it is likely to return to the original data island and application island.
So how to really achieve business-driven data center construction?
1. Global business insights
Identify new business opportunities and product opportunities by analyzing market behavior;
2. Personalized service
Personalized service refers to providing targeted products and services through accurate analysis of customer needs.
3. Real-time data reports
For business departments, what they want to know most after the launch of any product is the market’s feedback on the product.
4. Develop new business by sharing capabilities
The purpose of the Data hub is to abstract, share, and reuse data capabilities, where sharing and reuse are not only proposed to save money, but in many cases are the driving force for new business development.