[Huawei self-developed] | The rise of domestic database GaussDB

GaussDB

GaussDB adopts an integrated architecture and supports both relational and non-relational database engines, which can meet the comprehensive data intelligent management needs of government and enterprises, and realize high-speed operation and integration of databases. It is an important tool for intelligent upgrading of government and enterprises, bringing a new era of database management.

Introduction to GaussDB

GaussDB is an enterprise-level distributed database independently developed by Huawei. It has undergone long-term research and optimization, and has rich experience in telecommunications, finance, government and enterprise, and cloud technology. This database mainly addresses customers' needs for high availability, high performance, security and reliability, and is suitable for cloud environments.

GaussDB is a product series. In terms of overall architecture design, the bottom layer is distributed storage, the middle layer is the unique data structure of each DB, and the outermost layer is the interface of each ecology, reflecting the multi-mode design concept. Specific products include:

Relational Database:

• GaussDB (for openGauss), a distributed database based on the openGauss ecosystem
• GaussDB (for MySQL), a distributed database based on the MySQL ecosystem, is 100% compatible with MySQL

Non-relational database:

• GaussDB (for Mongo), a document-based distributed database, is 100% compatible with the interface of MongoDB, which solves the protocol risk of open source, and uses 3 copies to solve the reliability of 9 copies
• The distributed database GaussDB (for Influx) based on timing protocol solves the capacity problem of open source master and backup
• GaussDB (for Cassandra) based on wide table Cassandra solves the serious problem of GC jitter in the Java solution, and the data is strongly consistent;
• Based on the KV interface GaussDB (for Redis), currently 100% compatible with Redis, and can be compatible with more KV protocols in the future, the data is strong and consistent, and the large capacity is 10T+, which solves the problem of disconnection between open source master and backup

Database tools and middleware:

• Data Migration - DRS (Data Replication Service, referred to as DRS);
• Application structure conversion - UGO (Database and Application Migration UGO, referred to as UGO);
• Distributed Database Middleware—DDM (Distributed Database Middleware, DDM for short)

Product advantages

GaussDB brings together global resources, 2000+ database/data warehouse/big data experts and professionals in advanced core engines, algorithms, performance, etc. from 7 regions around the world, and has continued strategic investment for more than 10 years. In the financial, government and enterprise market, GaussDB's local deployment (Huawei Cloud Stack solution) has won the first market share of domestic databases; in the pan-Internet market, the growth rate of public cloud is the first (source: IDC China relational database software market data tracking in the second half of 2020 report); at the same time, the global DBMS market share has entered the top ten (source: Gartner Market Share Analysis: Database Management Systems, Worldwide, 2020); achieved large-scale commercial use in 1500+ financial government enterprises and major Internet customers; and 100+ partners Establish cooperative relationships and share market opportunities; and actively invest in university cooperation and developer ecology, accumulatively empowering more than 150,000 developers.

GaussDB full-stack products and solutions allow customers

• Safe selection : independent ecology + open source ecology, SQL+NoSQL, distributed and unified architecture, excellent performance, safe and reliable
• Migration at ease : from online data migration to application structure conversion, migration is worry-free
• Worry-free ecology : support open source and independent dual ecology

GaussDB(for openGauss)

• Ultra-high availability
  Supports cross-computer rooms in the same city, remote locations, multi-active high availability in different locations, supports distributed strong consistency, and zero data loss.
• Free expansion
  GTM-Lite technology breaks the bottleneck of traditional distributed performance. Computing and storage can be expanded horizontally freely, supporting 1000+ expansion capabilities, and performance increases quasi-linearly with expansion.
• Excellent performance
  Distributed optimizer + Numa-Aware technology blessing, providing the ultimate performance. Supports PB-level mass storage, and has excellent performance under various enterprise-level loads such as TPC-C and TPC-H.
• Open Ecology
  GaussDB (for openGauss) open-sources the main and backup versions of the single machine to create an openGauss open ecology. Customers based on the openGauss ecology can avoid the risk of being locked by a closed-source database and maximize the protection of customer investment.
• Rich enterprise-level features
  support distributed storage procedures, distributed triggers, external tables, transparent encryption, fully encrypted state, tamper-proof, WDR diagnosis report, intelligent diagnosis, index recommendation and other rich enterprise-level features, which effectively improve customer development operations. dimension efficiency.
• Safe and reliable
  Based on 10+ years of accumulation in the database field, a new generation of enterprise-level distributed relational database is independently developed. The core code has 100% intellectual property rights, and all modules of the product have code-level problem repair and architecture optimization capabilities.

Applicable scenarios : Applied to key core systems in industries such as finance, telecommunications, and government, and high-performance scenarios

GaussDB(for openGauss):
https://www.huaweicloud.com/product/gaussdb.html

GaussDB(for MySQL)

• Powerful performance
  Using separation of computing and storage, log as data architecture, RDMA network, performance up to 7 times that of open source databases.
• Data Reliability
  Based on distributed architecture, elastically expandable virtual block storage service; with high data reliability and high I/O throughput.
• Elastic and easy to use:
  1 write and 15 reads, adding read-only instances at the minute level, upgrading and upgrading specifications.
• High Compatibility
  Fully compatible with MySQL, without sub-database and tables, and applications can be easily migrated to the cloud without modification.
• Efficient backup
  Using Log Stream technology, quickly back up and restore terabytes of data in minutes, support backup storage for up to 732 days, and support data recovery at any point within the backup retention period.
• Massive storage
  Based on Huawei's self-developed DFV distributed storage, with a capacity of up to 128 TB, it supports Serverless automatic scaling according to data capacity; storage is automatically sharded, eliminating the need for separate databases and tables.

Applicable scenarios : medium and large enterprise production systems (high performance, large data capacity), such as finance, Internet, etc.

Official GaussDB (for MySQL): https://www.huaweicloud.com/product/gaussdbformysql.html

GaussDB(for Cassandra)

• Safe and reliable
  Protect database and user privacy through multi-layer security protection systems such as VPC, subnet, security group, SSL encrypted transmission, and fine-grained authority control; support deployment across three availability zones, fast backup and recovery; distributed architecture, the highest N-1 node failure tolerance.
• Super read and write
  Open source performance is improved by 3 times, supports real-time writing of online data 7*24 hours, has high availability and read and write load balancing capabilities.
• Elastic expansion
  Based on the distributed architecture of computing and storage separation, it realizes minute-level computing node expansion and second-level storage expansion, and the expansion process does not interfere with online business.
• Convenient management
  The visual instance management platform realizes one-click convenient operations for instance creation and deletion, backup and recovery, monitoring and alarming, and node expansion and contraction.

Applicable scenarios : Pan-Internet log data storage (large amount of concurrent writing, high storage capacity), industrial Internet data storage (large writing scale, large storage capacity)

GaussDB (for Cassandra) link:
https://www.huaweicloud.com/product/gaussdbforcassandra.html?modb

GaussDB(for Mongo)

• High performance
  Open source 3 times performance improvement, 7*24 hours online data real-time writing.
• Flexibility and elasticity
  Minute-level computing node expansion and second-level storage expansion can easily cope with changes in business traffic.
• Extreme reliability
  Multi-layer security protection, deployment across multiple availability zones, fast backup and data recovery.
• Visual management
  One-click convenient operations such as instance management, monitoring and alarming, and node expansion and contraction.

Applicable scenarios : Applied to games (equipment, props), pan-internet scenarios

GaussDB (for Mongo) link: https://www.huaweicloud.com/product/gaussdbformongo.html?modb

GaussDB(for Redis)

• High reliability
  Three copies of data persistent storage, strong consistency guarantee, stable and reliable performance.
• Low cost
  Save 75% of the cost and calmly handle various large data volume scenarios.
• Security operation and maintenance
  Provides operation and maintenance capabilities such as strict access control, real-time monitoring, and fault warning.
• Elastic Scaling
  Computing and storage resources can quickly and smoothly expand capacity to meet business peaks at any time.

Applicable scenarios : Key-Value storage mode, user Internet scenarios

https://www.huaweicloud.com/product/gaussdbforredis.html?modb

GaussDB(for Influx)

• High-performance writing
  Data is broken up according to two layers of "time Range + timeline Hash", distributed parallel writing, and can process trillion-level time-point writing at most every day.
• Flexible and Elastic
  Computing independent on-demand expansion, expansion without data migration, and the expansion and contraction of cluster nodes can be completed in minutes.
• High compression rate
  Columnar storage layout and dedicated compression algorithm, compared with the open source version, the compression rate can be increased by about 5 to 10 times.
• High query performance
  Multi-node multi-thread parallel query can efficiently handle high-concurrency and large-scale data analysis tasks.

Applicable scenarios : Industrial Internet time series data, user bank flow data, IoT data storage (time series)