background
As the youngest provincial power grid enterprise under the state grid, jibei electric power connects northeast China and north China power grids, and undertakes the important mission of power supply in beijing-tianjin-hebei region. In 2016, State Grid Hebei Electric Power Company undertook the promotion task of State Grid Corporation’s big data platform, initially built the enterprise-level big data platform, and comprehensively promoted the construction of public data resource pool to provide favorable conditions for centralized data sharing, analysis and utilization.
In order to speed up the construction of the global energy Internet and the goal of building a “one strong three excellent” modern company, in 2017, Jibei Company promoted and implemented the construction of the full-service data center according to the requirements of the State Grid. At present, it has access to 17 sets of information systems, and collected a large amount of data. In 2019, State Grid Jibei Electric Power Co., Ltd. started the construction of “Cloud Network Visualization and Analysis System Development” project.
In just a few years, Hebei Electric Power has completed the construction of four data centers, achieved unified data sharing and business integration, and accumulated a large amount of production operation and management data. However, there are shortcomings in data use and data analysis, and the methods of manual collection and manual analysis are often used to carry out research work on power grid planning.
The main challenge
In order to adapt to the new situation of energy supply and demand analysis and prediction, fully consider the impact of economic and social factors on the energy demand in northern Hebei, comprehensively utilize the multi-dimensional massive data, develop the energy supply and demand analysis and prediction platform, realize the functions of basic database management, regional energy economic analysis and power demand prediction, etc. And in the performance and reliability, information security, application and operation monitoring, maintainability, ease of use and system disaster preparedness to meet the relevant requirements.
Inside the hebei north power “unified national power market technical support platform, measuring total life system, power substation and team integration, electricity trading system, marketing business application, marketing analysis and auxiliary decision-making” flow of the business such as the rapid increase, the network operational complexity increases gradually, fault location and analysis of power demand is higher and higher. The development and construction of cloud network visualization and analysis system is a necessary supplement to the operation and maintenance work of Hebei Electric Power, which helps the operation and maintenance personnel to quickly locate and solve the risks and failures in the virtual network environment, improve the efficiency of cloud network operation and maintenance and reduce the operation and maintenance time cost.
The construction of cloud network visualization and analysis system of Hebei Electric Power Company faces the challenge of multi-location and multi-center at first. The second is the work of combing the complex service network in the cloud. Finally, the docking problem of automatic operation and maintenance system. In addition, in the project construction should adhere to the principle of “unified leadership, unified planning, unified standards, unified implementation”, and fully consider and use the existing data integration management and analysis platform of Hebei Electric Power, unified planning.
To deploy
According to the resource pool of Hebei Electric Power Co., Ltd., the deployment plan starts from the internal network monitoring of the data center in North Hebei, including the monitoring of the host side (virtual network) and the physical network (multiple network nodes), and realizes the multi-point monitoring of the main and backup links and the physical network. The system environment of Hebei Electric Power is a multi-site data center, and two data centers are selected for deployment in the first phase. The deployment diagram is shown below.
In view of the complex business in the cloud of Hebei Electric Power Company, Deepflow has sorted out the complex business by using the ability of docking multi-cloud platform, so as to quickly solve the monitoring problem of business network.
First of all, the cloud business of Hebei Electric Power was sorted out by IP, and dozens of B-class network segments were confirmed through backtracking and packet capture. After further confirming address information with network department staff, the information of about 150 C-class network segments was accurate.
Secondly, the physical network is depicted, the resource group topology is created, the business path is created, and the collection point is associated. On this basis, the communication model of the business network in the cloud is converged.
Thirdly, through the docking of automatic operation and maintenance platform and the manual import of ledger, the portrait of hundreds of businesses and thousands of paths is completed, and accordingly the process of business portrait is standardized. That is to create and sort out the resource group topology, draw the critical path through traffic statistics, compare and modify the business network information manually.
Through the sorting of IP resources, the order of magnitude of monitoring IP is greatly reduced. Through the docking cloud platform and communication modeling, the basic portrait of the service network is preliminarily sketched. On the basis of the first two steps, manually calibrating the business network finally realizes the precise monitoring of the whole network business.
The client’s return
Through the deployment of DeepFlow, Hebei Electric Power complements the ability to obtain comprehensive infrastructure monitoring data. Based on the unified data center of the whole service, the self-service analysis service component of the big data analysis platform realizes the visual analysis of the cloud network. It meets the demand of energy supply and demand analysis and prediction of the new situation, and contributes to the economic and social development and clean energy development of the power supply and service areas in Beijing, Tianjin and Hebei.
About DeepFlow ®
Deepflow ® is a virtual network traffic collection, visualization and analysis product launched by Beijing Spruce Century Network Technology Co., Ltd. (hereinafter referred to as Spruce Network). The Deepflow ® software is based on x86 clusters and common network devices and supports cluster deployment, container (Kubernetes), OpenStack, vSphere virtualization environments and public cloud environments such as AWS, Alibaba Cloud, and Tecum Cloud. It can help enterprises collect and distribute network traffic between arbitrary workloads in a hybrid cloud environment, realize comprehensive performance monitoring of cloud service critical links, and provide end-to-end full-path diagnosis of business network. With first-generation technology advantages and complete product solutions, Deepflow ® has been deployed in more than 50 Fortune 500 companies to continuously provide customers with best practices in cloud network governance.