Cutting-edge sharing｜Jiang Min, Co-founder & Vice President of Shulan Technology: Best Practices Based on Cloud Native Data Warehouse AnalyticDB PostgreSQL

Introduction to : 1619c67498af2e This content is for the 2021 Cloud Home Conference-Cloud Native Data Warehouse AnalyticDB Technology and Practice Summit Sub-forum, Jiang Min, the co-founder and vice president of "To share.

This article will introduce the best practices based on the cloud-native data warehouse AnalyticDB PG through four parts.

1. Background introduction

2. Program introduction

Three. Customer practice-a business group

4. Company introduction

1. Background introduction

At present, my country's data center industry is in a transitional period from budding to rapid development. With the explosive growth of data stars, advances in data processing technology, and the digital transformation of enterprises, the market demand continues to increase. The industry has an obvious growth momentum and the market scale is expanding rapidly. . At present, the data center has gradually become thinner in the industry, and for small and medium-sized enterprises with relatively simple requirements for data center capabilities, providing them with standardized and lightweight overall solutions will be the main demand of the market today.

When we build the capabilities of the entire data center, we will actually encounter some problems. For example, the underlying architecture is complex. Traditional data middle-office architecture thinking has led customers to select and deploy big data components, develop and schedule data tasks, and require multiple parties to weigh and consider the operation and maintenance and operation of data. The architecture is complex and the threshold for use is too high. Secondly, operation and maintenance costs are high. Accompanied by the complexity of the underlying architecture of the traditional data center, a complex IT personnel system and high operation and maintenance costs follow. Third, respond to the challenges of timeliness. With the continuous increase of data volume and the continuous iteration of business requirements, how to quickly respond to the needs of business personnel, reduce the waiting time for data analysis, and let business personnel have a better experience will become the key indicators for the success of the data center.

Two, program introduction

When we build the entire lightweight data center based on this version of ADB PG, you will find that there is an ADB PG underneath. In the middle, we are a lightweight data center suite. In this way, the two systems can support our upper-level business construction.

From the overall business architecture, the bottom layer is our existing system of various data sources. Then we will build a technical system in the data center, including ADB in the technical system. Then build our entire data system and service system. In the technical system of the data center, it is our data flow. It will go through a process of data modeling. For lightweight small, medium and micro enterprises, when a lightweight solution is needed, we will put it in ADB. In this process, there is no need to consider the problem of relocating data. All my data is in this system. The data volume of ADB can actually support the data usage of most enterprises.

As long as the enterprise builds the entire data model here, we will have corresponding tools to support it, including data synchronization, environmental management, model development, model monitoring, including some functions of data governance. For data and services, when we build the data model, we will build some specific indicators for the objects of the enterprise, and the corresponding insider system will also be built. For example, we build a customer's portrait system from the customer's perspective. From the supplier's perspective, build a supplier's portrait system. After the data is constructed, we can do insights, marketing, precise matching, etc. So from a data perspective, it can support application scenarios.

For specific scenarios, we are divided into several large blocks. The first one is the financial field. Financial statements are sometimes aligned with the corporate operating statements. When most companies face financial audits, it is a very strong demand for companies. The second is the management of the company's receivables and human resources. The insight of employees also requires digital means to manage employees called digital employees. More refined to do some assessments of employees, including some corresponding index analysis. Third, we are currently doing real-time data analysis in some areas. According to the trajectory to do situation analysis, including some collision analysis. Fourth, the ability to encrypt the database. This kind of scene can be built based on the ADB base. The availability of data is invisible, allowing us to manage data more efficiently.

3. Customer practice—a business group

We mainly talk about the case of a business group. This commercial group mainly manages high-speed service areas. In the original service area, each service area built its own system. Now that the concept of intelligent service area is proposed, the management of entering and leaving the service area needs to be unified. At the beginning of construction, we also encountered some problems. For example, data collection is difficult. 50 pairs of service areas are independent and can only collect part of the statistical data; a variety of information systems are built independently, and the management is not unified; the data of the surrounding road sections cannot be effectively obtained; the problem of data islands is obvious. The data management is chaotic, the data management departments are different, and there is no unified asset catalog; the data standards are different, and the data quality is uneven; the cross-departmental data use coordination is cumbersome. Data usage is weak, and most of the data is in a state of saliva. There is a lack of necessary intelligent insights for managers' operation monitoring, business decision-making, etc.

When solving these problems, consider how to reduce maintenance costs for this site in each service area? How can it be used out of the box? In fact, the construction scene in it has been briefly introduced just now. The first one is that there will be many scenes of video analysis in the server. When it comes to key information, it must be captured in real time and written to the Dao database. For example, the flow of people, the flow of vehicles, and the situation of congestion. After the people come in, what kind of products are there in demand? For the overall management and control of the service area, we are predicting the traffic in the service area, some supply-demand relations, and some management time. It needs to gather data to the headquarters for unified management and unified analysis.

From the perspective of customer value presentation, first, intelligence. Through the collection, processing, management and application of various data, combined with mature business models, we release the value of data and support business intelligence. Second, lightweight. We simplify the traditional middle-office architecture, replace the Hadoop ecosystem with ADB, reduce the complexity and cost of enterprise resource storage and calculation; provide guarantee for multi-dimensional analysis and decision-making. Third, it is scene-oriented. We take the specific business field as the most entry point, from bottom to top, carry out data governance around scenarios, build data assets, support data application requirements, and get close to value.

Four, company introduction

Shulan Technology was established in June 2016. Since its establishment, it has adhered to the mission of "making data available" to help enterprises continue to capitalize data and empower business innovation. Shulan Technology is headquartered in Hangzhou, with localized teams in North China (Beijing), South China (Shenzhen), West China (Chengdu), and Central China (Wuhan). The core members are from Alibaba, Huawei, IBM, Baidu, Kingdee, Oracle, etc. First-tier companies are among the earliest batch of big data service innovation practitioners in China. Up to now, Shulan Technology has used its core product "Shuqi Platform" to provide data center construction services for 1000+ enterprise customers, helping customers to use data center as a key digital technology to promote the digital transformation of enterprises.

Copyright Notice: content of this article is contributed spontaneously by Alibaba Cloud real-name registered users. The copyright belongs to the original author. The Alibaba Cloud Developer Community does not own its copyright and does not assume corresponding legal responsibilities. For specific rules, please refer to the "Alibaba Cloud Developer Community User Service Agreement" and the "Alibaba Cloud Developer Community Intellectual Property Protection Guidelines". If you find suspected plagiarism in this community, fill in the infringement complaint form to report it. Once verified, the community will immediately delete the suspected infringing content.