Shulou(Shulou.com)06/01 Report--

How short is a second? In the blink of an eye to read this article, it has been gone for a second, what can be done? On Singles Day in 2019, Aliyun's self-developed cloud native database POLARDB set a new record of 87 million database processing (TPS) peaks per second. 87 million calculations were completed in the blink of an eye.

At the technical media conference of Ali Yun double 11, Cao Wei, general manager of Aliyun Intelligent Database Division, said that with the joint support of Ali's self-developed database product POLARDB and a number of other database products, the whole system did not jitter 15 minutes before the start of Singles Day this year, ensuring that hundreds of millions of users shop smoothly. The peak TPS of POLARDB on Singles Day is as high as 87 million beats per second, far exceeding the traditional Oracle database.

During the double 11 period, Alibaba's core e-commerce system, transactions, inventory links, as well as e-commerce cloud, logistics cloud, financial cloud and other core applications all adopted POLARDB (All in POLARDB), providing users with a "silky and smooth" shopping experience, from platform to merchant, from logistics to customers. As Alibaba's independent research and development, China's first cloud native relational database, POLARDB is born on the cloud and serves the world. Adhering to the concept of "innovation, openness and inclusive benefits", through a series of innovative technologies, such as storage and computing separation, distributed shared storage, remote multi-activity, software-hardware integration and so on, it achieves minute-level resilience and millisecond physical replication delay, and easily cope with the massive order growth during the great promotion period. Bear the peak of trading volume which soared 122 times in an instant at zero point. 1. Distributed expansion and disaster recovery: the storage capacity can reach 100TB, and a single database can be expanded to 16 nodes to meet the storage needs of enterprise big data. Ensure that the inventory of goods is not oversold or overbought in the case of high concurrency. 2, remote multi-live global synchronization: RPO=0 database high reliability scheme, data reliability across AZ six copies as high as 12 9. Remote read-only nodes are synchronized efficiently, the kernel layer supports global disaster recovery, and strong hot line update capabilities ensure a smooth experience under the flood peak of zero traffic. 3. High performance, high flexibility and low cost: minute-level flexibility, which can cope with the business peak with the lowest cost. The flexible upgrade capability enables customers to schedule resources flexibly, ensure the smooth operation of Singles Day, effectively reduce the cost input, and meet the cloud needs of enterprise users in large-scale business scenarios. Rapid capacity expansion during peak periods to smoothly cope with business pressure; rapid downgrade during low peak periods to minimize business costs.

In addition to being 100% compatible with MySQL and PostgreSQL, the two most popular open source database ecologies, POLARDB is also highly compatible with Oracle syntax, providing a solution for traditional enterprise clouds that is less expensive than traditional database 1ax 6. Users can migrate to POLARDB with one click without modifying the application code, experiencing greater capacity, higher performance, lower cost, and more flexibility.

At present, POLARDB is the fastest growing database product of Aliyun, supporting not only business scenarios such as Taobao, Tmall and Cainiao of Alibaba Group, but also widely used in government affairs, Internet finance, new retail, transportation, education, games, social live streaming and other industries, serving many enterprises such as one trip in Zhejiang Province, Beijing bus, Yintai Department Store, China Post, Zhongan Insurance and so on. In the large-scale project of "up to one run" in Zhejiang Province, which benefits people's livelihood, POLARDB has opened up data isolated islands in different departments, so that data can be circulated in the cloud, convenient and safe. Beijing bus, the largest urban public transport system with the largest scale, the largest number of vehicles, the most complex model and billing system, uses POLARDB distributed database at the bottom, supporting more than 20, 000 buses in Beijing, with an average of 8 million passengers a day. POLARDB's leading security, reliability and high performance perfectly meet the requirements of the database in the field of finance and insurance. POLARDB enables Zhongan Insurance to increase the overall policy data reading and writing efficiency by 25%, realizing the sharing of data among multiple business departments and subsidiaries, while meeting the needs of high-frequency business development and financial-level scenarios. Yintai Department Store uses Aliyun POLARDB and other products to completely replace the traditional database, saving more than 10 million in IT facilities and database alone, and successfully cope with daily traffic peaks such as big promotion. Aliyun has the richest cloud database products in China, and more than 400000 databases have been migrated to Aliyun, covering leading enterprises in finance, telecommunications, manufacturing, logistics and other fields. According to Gartner, Aliyun already ranks third in the global cloud database market share, with an annual growth rate of 115%. Cloud native database, represented by POLARDB, is replacing the traditional database and becoming the first choice for enterprises. POLARDB has not only helped enterprises to solve the key problems of cloud in their core business, but also has been in the forefront of technological innovation. Relevant research results have been selected into international top database academic conferences (SIGMOD, VLDB, ICDE, etc.), and their technical strength has been recognized by authority. Papers related to ACM SIGMOD,POLARDB were selected this year. This paper describes the self-developed storage engine X-Engine designed for large-scale traffic scenarios. The innovation lies in breaking through the limitations of the traditional data architecture and adopting a new concept of hierarchical storage, which can reasonably relocate the data according to the data access frequency and realize the cache cache. In addition, "TcpRT: real-time quality of service acquisition and diagnosis system for large-scale massive cloud databases", "PolarFS: ultra-low latency file system for cloud databases" and "S3: a Scalable In-memory Skip-List Index for Key-Value Store" were also published on SIGMOD and VLDB respectively. In the database evaluation report "The Forrester Wave ™: Database-As-A-Service,Q2 2019" issued by the international authoritative organization Forrester, Aliyun Database has joined the camp of "strong performers". Last year, Aliyun also entered the Gartner database magic quadrant, the visionary quadrant, as the first and only Chinese company. At the sixth World Internet Congress a few days ago, POLARDB was elected the world's leading scientific and technological achievement of the Internet, standing on the C position with his strength.

The original link to this article is the original content of Yunqi community and may not be reproduced without permission.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.