Foreword:
In recent years, the concept of data center has been in flames, and there has been a wave of construction business center and data center in the market. So what is data center? What problems can it solve? First of all, the concept of data center was first proposed by Ali. Since Ali proposed the concept of "large center, small front desk", the concept of data center has become popular. Many companies, whether they are Internet companies or traditional companies, have built up. My own data center, what is the charm of the data center, so that companies can pay attention to it? This article mainly analyzes this problem from three aspects: what is the data center, how to do the data center, and why to do the data center.
What is the data center:
The broad data center includes data technology, such as a series of technology collection, calculation, storage, and processing of massive data. The data center includes data models, algorithm services, data products, data management, etc., and enterprise business It has strong relevance and is unique and reusable by enterprises, such as 1,000 basic models, 500 fusion models and 10,000 labels built by the enterprise itself. It is the precipitation of enterprise business and data. It can not only reduce repeated construction and reduce the cost of chimney-style collaboration, but also a competitive advantage.
How to build a data center:
(1) Data collection
Data aggregation is the entry point for data access to the data center. The data center itself does not generate data, and all data comes from business systems, logs, files, networks, etc. These data are scattered in different business systems and are difficult to use and generate business value. Data collection is the core tool that the data center must provide. The data of various heterogeneous networks and heterogeneous data sources can be easily collected into the data center for centralized storage to prepare for subsequent processing modeling. Data collection methods are generally divided into offline batch aggregation and real-time collection.
(2) Data asset management
Data asset management includes the management and display of data asset catalogs, metadata, data quality, data lineage, data life cycle, etc., presenting the enterprise's data assets in a more intuitive way and enhancing the enterprise's data awareness.
(3) Data service
The data service system is to turn data into a service capability, let the data participate in the business through the data service, and activate the entire data center. The data service system is the value of the data center. Enterprise data services are ever-changing, and the products in the Taiwan platform can carry some standard services, but it is difficult to meet the service demands of the enterprise. Most services still need to be quickly customized through the capabilities of the platform. The service module of the data center does not bring many services, but provides fast service generation capabilities and service management and control functions.
(4) Data application development
Data development is a set of tools for data processing and processing process control. The data development module is mainly for development and analysts, providing offline, real-time, algorithm development tools and task management, code release, operation and maintenance, monitoring, alarms and other columns. Integrated tools, easy to use and improve efficiency. According to the needs of the enterprise, a variety of data applications can be quickly built through a series of data tool sets at this layer, and provided to all parts of the enterprise in a timely and effective manner.
(5) Data security
Establish a sound and efficient early warning system, improve the monitoring mechanism of each module of the system, obtain quantitative data, analyze and find and predict potential crises, report in time, and warn of risks; provide data access black / white list design, such as setting up data access IP segment, if it is not in this IP segment, you will not be able to access the data assets; at the same time, based on the visual operation, you can flexibly control the deactivation and activation of the interface to achieve independent control of the calling process;
The benefits of building a platform:
The goal of the middle office is to improve efficiency, data operation, and better support business development and innovation. It is responsible for collaboration in multiple fields, multiple departments, and multiple systems. The platform is a natural evolution of platformization. This evolution brings a "decentralized" organization model, highlighting the ability to reuse capabilities, coordinate control, and differentiate the ability to build business innovation.
Should all companies build their own middle office:
Regarding whether all enterprises should build their own data center? In my opinion, it ’s not that we ca n’t be in the middle of Taiwan for the purpose of establishing a platform. We need to combine the company ’s own development needs with the basic information construction situation. I think there are several necessary conditions for the construction of the platform:
(1) The enterprise information infrastructure is well established;
(2) The data assets have accumulated for many years, but the data is relatively scattered and difficult to share;
(3) The front-end business needs are changeable, and the demand for data dependence is relatively strong, but the data and indicators are overlapping. From a technical perspective, more work is required to freely build and build similar building blocks;
(4) The technical department cannot meet the changing business needs in a timely manner.
Tool support:
The original intention of the NBI data analysis platform ( http://nbi.easydatavis.com:8033) is to provide simple, easy-to-use, low-cost, and fast online big data visualization analysis products and services to help companies improve their data operation capabilities
(1) Data preparation:
(2) Data modeling
(3) Data analysis and data report production
(4) Release
NBI data analysis platform effect display
big data analysis software, data analysis platform, open source data analysis platform, echarts, d3.js, data visualization analysis software, large screen data analysis display for
more information, please refer to ( http: //nbi.easydatavis. com: 8033 )