Alibaba Cloud: Increase the self-research of NoSQL database software and hardware integration technology

Introduction: On August 25th, at the Alibaba Cloud NoSQL Database Summit co-hosted by Tianchi Platform and Alibaba Cloud Database Division, Alibaba Cloud announced the NoSQL database self-development 2.0 plan to further increase the self-research efforts of the software and hardware integration technology system. By focusing on the three major directions of software-hardware collaboration, multi-modal integration, and cloud-native, the core competitiveness of NoSQL database products such as Tair, a cloud-native in-memory database, and Lindorm, a cloud-native multimodal database, has been enhanced.

On August 25, at the Alibaba Cloud NoSQL Database Summit co-hosted by Tianchi Platform and Alibaba Cloud Database Division, Alibaba Cloud announced the NoSQL database self-research 2.0 plan to further increase the self-research of the software and hardware integration technology system. Software-hardware synergy, multi-modal integration, and cloud-native integration enhance the core competitiveness of NoSQL database products such as Tair, a cloud-native in-memory database, and Lindorm, a cloud-native multimodal database.

NoSQL database generally refers to non-relational database (Not only SQL), which is widely used in the Internet, Internet of Things, Internet of Vehicles, financial securities, transportation and logistics and other fields. As the earliest domestic manufacturer to invest in the research and development of NoSQL database technology, Alibaba Cloud's NoSQL database has achieved industry-leading performance in a number of exclusive key technologies, forming a product matrix with cloud-native in-memory database Tair, cloud-native multi-modal database Lindorm, and Alibaba Cloud MongoDB as the core.

Market research firm IDC predicts that the total amount of data generated and stored in the world will reach 175ZB in 2025, of which more than 80% are unstructured data, including text, images, voice, video, web pages, etc. The explosive growth of semi-structured/unstructured and other diverse data will strongly promote the rapid growth of the NoSQL database market.

In order to further expand the leading edge of core products, Alibaba Cloud NoSQL Database announced that it will launch the self-developed 2.0 plan to innovate in Alibaba Cloud's next-generation cloud computing architecture such as high-performance computing, storage, network, and security, and increase core hardware capabilities. Self-research efforts focus on the three directions of software-hardware collaboration, multi-mode integration, and cloud-native, providing the ultimate cost-effectiveness with a completely cloud-native architecture, ensuring business stability and data security with comprehensive enterprise-level features, and supporting rich data processing functions Digital and intelligent upgrade of customer business.

Taking Tair, a cloud-native in-memory database as an example, after the Tair software is completely self-developed, it will be fully based on the self-developed cloud-native infrastructure, from the Panjiu server based on Yitian CPU, to the Feitian cloud computing operating system, to the self-developed AliSCM Storage-level memory components, and then self-developed CXL memory pooling, form a complete self-developed hardware and software integration technology system. Based on the self-developed infrastructure and through full-stack optimization, Tair will be comprehensively improved in terms of persistence capability, cost, resource elasticity, and complex computing, and will define a new generation of cloud-native in-memory databases.

At the meeting, Jobbang and Cathay Pacific Property & Casualty shared the practice of using Alibaba Cloud NoSQL database as typical customers. After introducing the Tair database, Jobbang used persistent memory as the storage medium, and the cost was reduced by nearly 40% while the performance was basically the same as that of full-memory Redis; Cathay Pacific Insurance introduced the Lindorm database, and in the historical policy analysis scenario, the query performance was improved. At the same time, through Lindorm's deeply optimized ZSTD compression algorithm, the storage efficiency is further improved by 30%, and the overall comprehensive cost is reduced by 75%.

At the same time, Alibaba Cloud's NoSQL database also announced the latest progress in 2022. Among them, Tair, a cloud-native in-memory database, launched a new high-performance enterprise version, a large-capacity disk version, a persistent memory version, and a graph engine version; the cloud-native multi-mode database Lindorm released a new version for the Internet of Things. , the new functions of the Internet of Vehicles, greatly improve the storage and access performance of massive data; Cloud MongoDB 6.0 will be officially launched in September, when you can experience new features such as time series collections supporting secondary indexes & compound indexes.

Li Feifei, head of database products at Alibaba Cloud, said that NoSQL is a must-have database in the era of the Internet of Everything. In the future, we will continue to increase our self-research efforts, continue to reduce the management cost of each byte of data, and help customers share the development opportunities of the digital economy.

According to reports, Alibaba Cloud NoSQL database was born out of Alibaba's e-commerce business's demand for low-cost and ultra-large-scale distributed data processing. In response to the business data of massive e-commerce platforms, Alibaba has started a long journey to develop its own database. During the Double Eleven in 2021, Tair will guarantee sub-millisecond latency under the peak QPS, and Lindorm supports hundreds of petabytes of online big data scenarios for all businesses. Millisecond access and extreme cost reduction in high-level storage. Alibaba Cloud NoSQL database now has a rich product family, which has great advantages in performance and cost performance.

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