Huawei cloud gaussdb heavily released HTAP for commercial use, defining a new paradigm of cloud native database 2.0

In recent days, , At Huawei partners and Developers Conference 2022, Hua Wei Yun CTO Zhangyuxin released 《 Cloud native 2.0 Architecture white paper 》, Including cloud native databases , Cloud primitives are introduced 2.0 Key features of 、 Architecture mode , And good practices , It injects cloud Nativity into enterprise digital upgrading 2.0 fresh impetus . Fengke, chief architect of Huawei cloud database, shared the cloud native database at the meeting HTAP Significant features are commercially available , Through the extreme mixed load capacity and timely and accurate data analysis , Help enterprises make business decisions .

Hua Wei Yun CTO Zhang Yuxin made a speech on cloud primordial 2.0 Important speech

Cloud native database technology Again innovation

As early as last year , Huawei cloud released 《 Cloud native 2.0 white paper 》, Cloud primordial 2.0 Definition 、 Value and reference framework , Read “ Resource efficient 、 Applying agile 、 Business intelligence 、 Safe and trustworthy ” Four values and prediction of future development trend . And this year's 《 Cloud native 2.0 Architecture white paper 》, Upgrade around architecture , It mainly interprets cloud primordial 2.0 Technical features and architectural design patterns , Promote enterprise digital transformation and upgrading with cloud native technology .

Cloud native database is one of the important contents of the white paper , Major breakthroughs in architecture innovation and technical features . Huawei cloud native database GaussDB Based on the separation of storage and computation , And use cloud native database cloud storage DFV Based on the semantic pushdown ability of storage operators , Support parallel push down of transaction operators ; At the same time, the storage layer supports log playback , The database write node only needs to write logs to the storage layer , The storage layer can play back the logs as data pages , And provide consistent versions on multiple copies , It greatly saves the high-speed network bandwidth of the computing layer and the storage layer . Optimized cloud native GaussDB The architecture can well solve the hot data problem , And data loss 、 Problems such as fault recovery , Facing the high load of customers and various complex scenarios , Behave more calmly .

Serverless It is also one of the key features of cloud native database . Huawei cloud native database GaussDB combination Serverless, Just create the database node , Specify the required database capacity range , Then the database is charged according to the database capacity used per second during the working state of the database , At the same time, Huawei cloud database application lossless transparent switching technology (ALT), The specification change can make the application basically imperceptible , Let developers focus on application development , No need to focus on resources .

Besides , Fengke, chief architect of Huawei cloud database, also introduced the cloud native database GaussDB Another important feature of ——HTAP. Cloud native HTAP Deep integration OLTP and OLAP The advantages of both , Unified entrance , Automatic shunt , And establish a globally consistent transaction view on the row and column storage engines , Achieve real-time data consistency , Scenarios that support more mixed loads . This method can support a large number of concurrent updates , And the data synchronization time is delayed to seconds or milliseconds , It effectively avoids data extraction in traditional solutions 、 Cumbersome steps such as conversion and loading , It greatly improves the timeliness of data processing , Improve real-time and accurate decision support for enterprises . At present, this feature has been officially commercialized .

Cloud native database Good practice

Cloud native database GaussDB Not only continuous innovation in technology , It is also widely used in e-commerce 、 manufacturing 、 traffic 、 Pan Internet and other fields , Through the tempering of key industries , And achieved significant practical results .

In the e-commerce industry , In response to the development needs of mengfen group , Huawei cloud native database GaussDB(for MySQL) Help mengguan group build a high-performance cloud native e-commerce platform , Through smooth expansion and fast elastic lifting capacity , It can deal with sudden traffic calmly , After the big promotion, it can also fall back in time , Fast cost reduction . Moreover, the operation and maintenance efficiency has been improved by about 30%, The number of orders per second reached a new record , It has become a benchmark for using digital infrastructure to boost the rapid growth of business .

In the automotive industry , Huawei cloud native database GaussDB(for MySQL) Success helps FAW red flag ERP System microservice transformation and upgrading ,ERP The overall performance of the system has been greatly improved , Business operation is fast and stable under the flood peak ; Provides an open MySQL ecology , Business collaboration is more flexible . Use cost , Each database consists of the original minimum 600 More than ten thousand / Years down to 100 More than ten thousand / year ; In terms of production efficiency , It takes a few days to ship bom It is the fastest to arrive at a single destination 20 You can calculate the... Of a production plan in minutes bom single ; The splitting of microservices improves the stability of the system , Improved business delivery , The system is more scalable .

Cloud native database Future trends

The wave of cloud native technology is surging . In cloud native 2.0 after , How does the cloud native database ride the waves 、 Ride the wind and waves ？

With span Region Technology enhancements , Drive for high resource utilization , as well as Serverless The universal appeal of , Go to Region turn 、 Smart resilience 、Serverless Database will be the future development direction .

Before that , Cloud native database still has a way to go , such as Serverless In the future, databases need to be intelligent and resilient , It can calculate the user portrait according to the user's historical load , Rely on second level monitoring capability , Quickly judge the load trend of users in the future , Stretch resources for users in advance , Avoid the impact of user load on the upper limit of resource specification , Reduce system resource waste .

Although there are certain challenges , But it's also very dynamic , Huawei cloud native database GaussDB It will gather technical strength , Be creative , Help more enterprises upgrade their business intelligently , Deepen data service capability , Promote the digital transformation of a thousand lines of business .

Source of the article ：https://www.163.com/news/article/HAFT4N7J00019UD6.html