The real-time computing Flink version is a fully hosted serverless Flink cloud service provided by Alibaba Cloud. It has full lifecycle capabilities such as job development, data debugging, operation and monitoring, automatic tuning, and intelligent diagnosis of real-time applications.The kernel engine is 100% compatible with Apache Flink, with twice the performance improvement. It has enterprise class value-added functions such as FlinkCDC and dynamic CEP, and has built-in rich upstream and downstream connectors to help enterprises build efficient, stable and powerful real-time data applications.
Real time data warehouseAssist enterprises to establish a digital operation system, make real-time statistics of key business indicators, visually present the production and operation of enterprises, and provide real-time data support for business decisions
Real time ETLBuilt in integration of dozens of connectors, covering database, message queue, OLAP engine and other systems.Self written SQL uses full hosting services to integrate real-time data flow and help enterprises build data middle ground
Real time risk controlReal time identification of business risks such as scalping, cheating, malicious crawlers, etc. from massive data to avoid huge economic losses for enterprises. It can also use CEP (Complex Event Processing) to directly detect abnormalities in streaming jobs
Real time monitoringReal time computing Flink version's efficient status management, rich window support and other features can help enterprises simplify the rule alarm configuration process, improve the alarm effect, and reduce the cost of monitoring platform maintenance
Real time recommendationReal time analysis of user behavior, combined with AI technology to establish more accurate user profiles, and timely recommend more suitable news, videos and goods to users
Real time IoT data analysisCapture and analyze the huge amount of data generated by IoT equipment in real time, help users to analyze and diagnose the operating condition of equipment in real time, detect operating faults in real time, and predict the yield of products in real time
Product advantages
Full custody Flink service
Out of the box Full cycle of development, operation and maintenance Flexible billing
Rich enterprise level capabilities
One stop development, operation and maintenance platform integrating streaming and batch Flink CDC real-time warehousing Dynamic CEP helps real-time risk control/marketing Automatic job tuning, making full use of resources, intelligent diagnosis and rapid positioning
Kernel engine optimization: CPU has hundreds of thousands of records per second The test performance of Nexmark is 200% higher than that of open source State storage backend optimization
100% compatible with open source
100% compatible with Apache Flink Support open source Flink smooth migration to the cloud Seamless connection with mainstream open source big data ecology
Custom connector, UXF capability extension Open API helps users integrate their own systems User development capability expansion and precipitation reuse User's own system integration
Industry recognition
Authoritative certification of China Academy of Information and Communication The only real-time stream computing product in China entering the Forrester quadrant Financial real-time digital warehouse scheme was shortlisted in the typical directory of information innovation of the Ministry of Industry and Information Technology
Operation and maintenance services of the original factory are fully covered, and the SLA is 99.9% guaranteed
Full link automatic fault tolerance, supporting JobManager fault tolerance
Provide single point fast fault tolerance, flexibly balance data consistency and availability
Provide complete system checkpoint and job snapshot lifecycle management
Provide status compatibility check and status data migration
Support tenant level and project level resource and code isolation to meet cross team collaboration requirements
Connect with Alibaba Cloud account system and support multi role access control
Real time calculation Flink version (out of the box)
Initial purchase of new users 0 ¥
Open source Apache Flink
Industry solutions
E-commerce industry
Game industry
Financial industry
Advertising industry
Internet transactions
With the rapid development of the domestic e-commerce industry, the original "wide range" and "wide spread" operation mode can no longer meet the business needs.Through real-time computing, we can quickly mine user characteristics, analyze demand preferences, and help e-commerce enterprises complete the transformation of digital, refined, and personalized operations.
Capable of providing
Real time user profile
Develop real-time, efficient and personalized recommendations for different types of customers to improve the conversion rate of orders
Real time association recommendation
Establish real-time correlation analysis after customers purchase information to improve the probability of second purchase
As an emerging entertainment industry, the game's ability to attract money and interest has attracted more enterprises and talents to invest in it.With the rise of app stores and social networks, the scale of the game market has expanded unprecedentedly. Big data and AI have played an increasingly significant role in game operations, especially in extending the life of games.
Capable of providing
Forum data tracking
According to the communication and control of the player forum, access to the current player's core concern information and game bug detection to ensure the fairness of the game
Operation data control
Through real-time computing technology, conduct real-time analysis on game operation activities to obtain the most effective user transformation channel and improve the operation and promotion effect
Behavior data analysis
Establish player user profiles based on real-time behavior data analysis of players, customize different game benefit modes, and improve player retention rate and conversion rate
In recent years, with the rapid development of the Internet financial industry, the combination of various businesses of traditional financial institutions (such as exchanges, securities companies, banks, etc.) with the Internet has become increasingly close.Provide overall solutions for the demand link, risk control, credit assessment, fraud risk, data quality and other needs of banking business.Take financial institutions as an example, build real-time data warehouse and anti fraud system through real-time computing Flink version, and help financial institutions quickly build real-time risk control system.
Capable of providing
Fraud detection
Real time monitoring and differentiation shall be carried out for user accounts and transaction behaviors to quickly identify transaction information and reduce user losses;
Real time risk control
Optimize the traditional risk control system through real-time calculation of Flink version, transform intelligent risk control, and improve the overall identification efficiency to apply multiple scenarios;
Credit evaluation
Quickly identify malicious fraud such as false data and information counterfeiting, identify high-risk groups, and shorten the audit process through intelligent ways;
The development of mobile Internet has gradually reduced the threshold and cost of obtaining information services, and also made users' information consumption more frequent and diversified.The overall growth rate of the mobile network news industry market is fast, and the development potential is huge. With the substantial growth of mobile traffic, the number of mobile news information users increases, and the differentiation is significant. The refined operation will be the main theme in the future.
Capable of providing
Massive data processing
Use powerful data processing capability to cope with the growing data volume (the number of users of the platform, daily average UV/PV and other core indicators), and improve business efficiency.
Standardization of complex scenes
Mainstream media, such as news, short video, live broadcast and other output methods, establish standardized and universal scenarios, and make business management and control faster.
Fast real-time business output
According to the personalized business needs of the content sharing platform, the business is refined through real-time computing to increase user stickiness and consumption desire.
Real time content hotspot analysis, real-time user behavior analysis, real-time monitoring and analysis of user growth and activity, real-time recommendation of article videos, real-time advertising, real-time traffic operation, etc. are carried out through real-time computing Flink.
Online education enterprises
By using Alibaba Cloud's real-time computing Flink version, you can effectively solve the real-time/accuracy problems of computing in key online education scenarios such as continued reporting, resumption of classes, and live video quality monitoring.The complete streaming SQL capability provided by the product can greatly reduce the cost of business access and development. At the same time, the automatic elastic expansion and contraction capability of the product can effectively cope with traffic peaks and better meet the needs of complex and changing scenarios.
Online trading enterprises
By using Alibaba Cloud real-time computing Flink products, users can build the entire real-time recommendation system to improve the retention rate and conversion rate;At the same time, the landing of a large number of labels and customer portraits also enables the operation team to conduct more accurate targeted promotion for customers, significantly improving the conversion rate of promotion content.
Financial enterprises
Real time computing Flink version is used to build a real-time digital warehouse and real-time anti fraud system, helping financial institutions quickly build a real-time risk control system.The main business scenarios include real-time anti fraud, real-time large screen, real-time accurate push, etc.
Customer Stories
Aunt Qian
The risk control business architecture of Qian Dama is mainly composed of four parts: event access, risk response, risk perception, and risk backtracking. Alibaba Cloud real-time computing Flink version product online ETL processes real-time user profile tags and sales fact indicators, providing important data source support for the event access of Qian Dama's real-time rule engine.The risk control technology scheme has been applied to the business of Qian Dama Group to accurately identify the risk points of marketing activities and protect the interests of users.
Yicang Technology
ECom Technology is committed to building an intelligent and collaborative cross-border network. At present, tenants are distributed around the world, and the data comes from all continents. It collects sales data from all sites of major overseas e-commerce platforms.Faced with such a large data volume and complex scenarios across all regions of the world, the big data team of E-warehouse uses the ability of real-time computing Flink version of the high concurrency pipeline to process data and achieve accurate millisecond level synchronous warehousing of data.
Jiangling Motors
As a top 20 automobile manufacturer in China and one of the Fortune 500 Chinese enterprises, Jiangling Automobile Co., Ltd. has chosen to cooperate with Alibaba Cloud in big data construction. Through Alibaba Cloud real-time computing platform, it has built a unified real-time computing center for the company, serving multiple real-time scenarios of three business lines, namely, vehicle center, smart fleet, and real-time big screen,It realized the second level response of real-time business, and successfully helped the business department to improve the decision-making efficiency.
Chanjet
Changjietong's "Easy Accounting", as an accounting tool for the integration of industry, finance and tax, requires the administrator to query the statistical summary information of the tax declaration and voucher carry forward of the enterprise under his jurisdiction in real time.After using the real-time computing Flink version of the product, every tax declaration and carry forward is pre calculated in memory, and the query results can be displayed in real time in front of users, which greatly improves the response speed and user experience.
Two-dimensional fire
2D Fire focuses on the R&D and application of cloud computing catering software system.Its big data team realized high concurrency and second level analysis and processing of 10 billion level data through Alibaba Cloud real-time computing Flink+Hologres, and provided customers with more than 80 different types of reports, including business, order, dishes, members and other reports, providing multi-dimensional analysis perspective and business exploration for store development.
Weimiao
Through real-time calculation of Flink products, Weimiao has built a perfect real-time digital warehouse storage system to support businesses such as real-time monitoring of live broadcast indicators, improvement of class arrival rate, real-time monitoring of landing page access, effectively improving the class arrival rate by 10.5%, renewal rate by 1.5%, and stably supporting business needs.
Palm Reading Technology
Palm Reading Technology focuses on digital reading and is one of the world's leading digital reading platforms.Massive users based on the digital reading platform will build data infrastructure platforms such as commercialization, user growth, and recommendation services through Alibaba Cloud real-time computing Flink and other big data computing and analysis services to achieve the combination of commercial value-added and user reading experience.
Flying crane
Feihe is one of the earliest milk powder enterprises in China. It has a forward-looking vision to layout exclusive industrial clusters and control the whole process of milk powder production.With the help of real-time computing Flink and other products, Feihe builds a digital operation management system, establishes a real-time early warning mechanism, spies on risks in advance, and improves the effectiveness of supply chain and other business decisions.
Product Dynamics
2016-11-15 New functions/specifications
Stream computing supports log service read/write function
View details
2016-12-15 New Features/Specifications
Stream computing supports RDS table dimension function online
View details
2016-12-15 New Features/Specifications
Stream computing supports the display of performance and resource indicators of job dimension CPU and MEM dimensions
