Jiangsu Mobile's New Generation Business Support System Realizes the 1 2 X Mode and Improves the S

Recently, Jiangsu Mobile introduced new stream native technology, combined with cloud native technology system, and completed the provincial launch of a new generation of business support system based on stream cloud integrated architecture, realizing the new evolution of the support system in the cloud native era. Facing the new challenges of the 5g era, the new generation of business support system has created a new support architecture by introducing new ideas, new technologies and new methods; Through cross system resource integration, capability intelligence and data integration, combined with micro service and AI capabilities, the supporting objectives of large-scale, agility, intelligence and flexibility are realized; By creating a new generation of efficient and intelligent streaming cloud integrated infrastructure, it can effectively help improve the company's it support efficiency, which is mainly reflected in the following aspects:

·It has full flexibility, realizes flexible scheduling and sharing of resources, and continuously reduces resource costs

Introduce stream native technology to realize the elastic expansion of stream processing applications, and combine the elastic expansion of transaction oriented micro service applications realized by cloud native technology to comprehensively expand the range of system elastic capacity, improve system processing capacity and reduce resource costs.

·Unify the technology stack and processing framework to realize the 1 2 X support mode and improve the support efficiency

Based on the stream native technology, build a stream processing framework with pulsar as the core to support the stream processing application of the full support system. Combined with the micro service framework with servicemesh as the core and the PAAS platform with container as the core in the cloud native system, realize the unified technology stack, unified framework and unified capability standard of the support system, form a 1 2 X support mode, and align the elastic expansion and expansion of each system Gray scale publishing, agile delivery, intelligent operation and maintenance, micro service management and flow processing management, so as to improve the support efficiency and architecture control ability.

·The support system is fully intelligent, realizes it replacement, and improves the intelligent level of the support system

In terms of intelligent resource scheduling, intelligent elastic scaling, performance optimization, intelligent traffic fragmentation, intelligent response to business consulting, root cause analysis and capacity prediction, artificial intelligence algorithms are fully introduced, combined with system operation and maintenance data to realize intelligent processing and prediction and improve support efficiency.

·Focus on customer perception, achieve key breakthroughs in key scenarios, and improve internal and external customer satisfaction

Through the introduction of stream native technology and combined with the ability of cloud native, reconstruct the scenario application architecture such as 5g bill processing, massive signaling processing, economic daily report, marketing / reminder / signal control / business handling event processing, so as to comprehensively improve the processing efficiency and enhance the perception of internal and external customers.

Figure 1 construction results

Architecture evolution background

Figure 2 current system support challenges

Jiangsu Mobile's X86 based third-generation service support system has been used for many years since its completion in 2014, with a significant increase in business volume and work orders. At the same time, a hundred flowers bloom in chbn business applications in 5g era, and the number of messages, bills and signaling data will increase more than 10 times. The support system is facing the following challenges:

·Massive data processing: in recent years, the average daily traffic of users in our province has increased from hundreds of TB to tens of thousands of TB. With the large-scale commercial use of 5g, the amount of messages, bills and signaling will continue to increase sharply, resulting in a significant increase in the processing pressure of the supporting system.

·Customers' perception requirements are higher: 5g users have double PCRF, and the number of work orders opened has doubled. It is required to continuously improve the work order processing efficiency and ensure the timely opening of services; 5g traffic is used faster, users are more sensitive to traffic costs, and the business department requires further acceleration.

·Continuous investment of hardware resources: with the increase of business volume, the needs of computing and storage resources are increasing to meet the needs of business development, and the cost of hardware resources is increasing.

In order to solve the above problems, after several rounds of expert evaluation, technology selection, test and verification, the stream native technology represented by pulsar is finally fully introduced, and combined with the cloud native PAAS platform, we can jointly realize the comprehensive evolution of the new generation of business support system.

Overall architecture scheme

Figure 3 cloud Integration Architecture

The construction core of the new generation business support system includes three aspects: stream native technology framework with pulsar as the core, micro service framework with servicemesh as the core, and cloud native PAAS platform with container as the core.

·Stream native framework: for stream processing applications in boss / CRM / big data systems, it implements unified framework encapsulation based on pulsar, and provides general capabilities such as distributed message queue, distributed computing engine, I / O connector, SQL Engine, load balancing, distributed hierarchical storage, multi tenant isolation, thermal distribution, link tracking and elastic scaling, so as to simplify the complexity of stream processing applications, Improve support efficiency.

·Microservice framework: facing the transaction application in boss / CRM / big data system, with servemesh as the core, it provides general capabilities such as balance, service discovery, service routing, fuse current limiting and monitoring measurement, unifies the service framework, improves the service governance capability of each system and improves the support efficiency.

·Cloud native PAAS platform: build a cloud native PAAS platform with container as the core, provide elastic scaling, elastic computing, automatic delivery, gray publishing, intelligent operation and maintenance and other capabilities, improve delivery efficiency and reduce operation and maintenance costs.

Key technical characteristics

Figure 4 characteristics of six technical capabilities

Liuyun integrated architecture has six capabilities: structured schema and SQL support, hierarchical and hierarchical storage, function computing features, storage and computing separation architecture, cross center replication, operation and maintenance and delivery capabilities. The current production environment has begun to take shape:

Figure 5 production cluster capacity

Key scene support

1. Create a new billing architecture, realize the "six modernizations" support of the billing system and improve the support efficiency

Figure 6 billing center based on native stream

Based on stream native technology, combined with micro service and AI capabilities, the billing core architecture is rebuilt to realize the support of "scale, messaging, micro service, agility, flexibility and intelligence" of the billing system, and greatly improve the processing efficiency. Through service-oriented, the number of landing times of bill messages is reduced by 8, the bill processing performance is improved by 180%, and the number of equipment is reduced by 26%. Through messaging, the interaction between billing system and Shared Storage NAS is reduced by 18 times, and the burden of NAS resource I / O is reduced by 90%. Through elastic scaling and intelligent scheduling, flexible scheduling and global sharing of resources in free and busy hours are realized, and the resource utilization rate is improved by 35%; Through gray-scale release and agile delivery, the on-line process pipeline delivery and bill processing are uninterrupted, so as to improve customer satisfaction.

2. Reconstruct the real-time processing of massive signaling, maintain the freshness of data and improve the value of data

Fig. 7 native real-time stream processing platform

Using stream native technology, reconstruct the real-time stream processing platform to realize stream processing. It has the capabilities of "elastic expansion, multi tenant isolation, hierarchical data storage and offline data analysis", so as to comprehensively improve the timeliness of signaling data processing, maintain data freshness and give full play to the maximum value of data. Through the separation function of pulsar storage and computing, the elastic expansion capacity is realized, the signaling processing capacity is increased by 4 times, and the real-time data processing capacity of 500 billion / day can be supported; Through multi tenant capability, realize resource and authority isolation between data, and improve the security of stream processing; Through hierarchical data storage, cold and hot data are stored separately, which improves the durability of data storage; Through seamless connection with the analysis engine, the offline analysis ability of signaling data can be realized, and the advanced nature of the architecture can be enhanced.

3. Build a real-time data warehouse, realize the overall speed-up of economic sub daily reports, and improve the timeliness of business decisions

Figure 8 real time report

Based on stream native technology, build a real-time data warehouse, realize the real-time transformation of data warehouse, and improve the real-time performance of economic express and daily report data. Through the capability of pulsar multi-I / O connector, realize the rapid access of multi-channel data and improve the efficiency of data access and change perception; Through the real-time processing of data assets, realize the real-time processing of daily and express data, help managers at all levels to timely control business development and improve the timeliness of business decisions.

4. Build an event processing center to achieve efficient processing of customer related events and improve customer perception

Figure 9 event processing center

Based on stream native technology, build a real-time event processing center to realize event sharing, subscription and efficient processing, and improve the processing efficiency and reuse of events. Through the event message processing, the timely rate of marketing, information control and reminder is improved from minute level to second level, so as to improve customer satisfaction; Through the message subscription mechanism, realize the reuse of information control, reminder, data change and other events, reduce data redundancy and improve data consistency.

Jiangsu Mobile's New Generation Business Support System Realizes the 1  2  X Mode and Improves the S 1

the support efficiency related articles
What Are Some Ways to Keep My Hair Really Healthy?
How Was the First Sale of Vivo X60 Series?
Huami Technology Held a Press Conference on August 27, and Amazfit Smart Sports Watch 3 Officially A
Connect to Nonencrypted Wireless Network Using Ubuntu Commands
Is Time Travel Possible? Can We Travel Back in Time?

Copyright © 2020 Coffee bag - Guangzhou tianci packaging industry Co,. Ltd. | Sitemap