Gold Award Solution | Versatile and World-Class: What Makes GBase 8c So Powerful?

Published on 2022-11-10

Kunpeng Application Innovation Contest is a top-tier competition for global developers. This contest is co-hosted by 21 Kunpeng Ecosystem Innovation Centers and Huawei, together with the China Software Industry Association, the Green Computing Industry Alliance, the China Computer Industry Association, and the High Performance Computing Committee of the China Computer Federation. It aims to stimulate industry application innovation, accelerate industrial integration, promote talent development, and attract developers across the industry to jointly create full-stack Kunpeng solutions.

Gold Award Solution in the National Finals openGauss Track

Tianjin General Data Technology Co., Ltd.:“Multi-Model and Multi-Modal Distributed Database Solution Based on openGauss”

 

Solution Introduction

GBase 8c is a multi-model and multi-modal distributed database built on openGauss 3.0. It supports multiple storage modes such as row-based, column-based, and in-memory, as well as various deployment forms including standalone, primary-standby, and distributed. GBase 8c features high performance, high availability, elastic scalability, and high security. It can be deployed on physical machines, virtual machines, containers, private clouds, and public clouds, providing secure, stable, and reliable data storage and management services for core systems in critical industries, internet business systems, and government and enterprise systems.

A High-Value, Highly Innovative Distributed Database

GBASE (General Data Technology) is one of the few independent database vendors in China that focuses on database product R&D and has achieved large-scale deployment in the financial and telecommunications industries. General Data Technology officially joined the openGauss community at the end of 2021, yet it is the first in the industry to launch a distributed database product based on openGauss 3.0. Developing a distributed database on openGauss requires a deep understanding of the openGauss architecture and design, and then building distributed capabilities on top of it. With nearly two decades of database R&D experience, the company's flagship product GBase 8a is a distributed analytical database. This expertise provided a solid foundation for developing GBase 8c, a native distributed database based on openGauss.

As the only native distributed database built on openGauss 3.0, GBase 8c has undergone extensive optimization at the system architecture and kernel levels, compared to solutions that provide distributed capabilities through middleware. In particular, it offers core advantages in strong data consistency and complex query performance.

GBase 8c delivers global strong consistency. By using global timestamps issued by a global transaction manager, it ensures strong consistency for all data within the entire cluster, making it more suitable for core business systems than the eventual consistency offered by middleware-based solutions.

GBase 8c supports data exchange between data nodes. For complex queries, middleware solutions can only pull all data to the middleware for filtering, which increases network load and middleware processing pressure with large data volumes. GBase 8c can push down complex queries to the data nodes for processing, where the data nodes perform data exchange and computation. Through parallel processing across multiple data nodes, the overall cluster processing efficiency is greatly improved.

These capabilities of GBase 8c allow users to use distributed databases just like centralized databases, enabling a smooth transition of database architectures.

Strengthening Distributed Capabilities, Delivering One-Stop Solutions

GBase 8c is a native multi-model and multi-modal distributed database built on openGauss 3.0, supporting multiple storage modes such as row-based, column-based, and in-memory, and various deployment forms including standalone, primary-standby, and distributed.

What is “multi-model and multi-modal”? “Multi-model” refers to multiple storage modes based on the row-based, column-based, and in-memory storage engines provided by openGauss. GBase 8c supports multiple storage engines to meet business needs in different scenarios. Specifically:

① Row-based storage engine:

Designed primarily for OLTP scenarios, such as order processing, shipping, and banking transaction systems;

② Column-based storage engine:

Designed primarily for OLAP scenarios, such as statistical reporting and analytical systems;

③ In-memory engine:

Designed for extreme performance scenarios, such as bank risk control.

“Multi-modal” refers to multiple deployment forms. Building on openGauss's original standalone and primary-standby deployment modes, GBase 8c adds distributed deployment capabilities, offering three deployment forms: standalone, primary-standby, and distributed. The standalone form is the most classic database deployment, with advantages of low cost and simple deployment. The primary-standby form supports a one-primary-multiple-standby configuration, offering simplicity, efficient delivery, and certain disaster recovery capabilities, suitable for scenarios with relatively small data volumes, extreme single-node performance requirements, and a need for high availability. The distributed form supports full component redundancy for high availability and separation of storage and compute. Its advantage lies in the independent horizontal scaling of compute and storage resources according to business needs, making it ideal for high-volume, high-concurrency scenarios demanding high data security.

Co-Creating Leading Database Products to Accelerate Core System Deployment in Industries

During the development of GBase 8c, the openGauss team provided substantial assistance. In the system conception phase, the openGauss team organized multiple training sessions for developers, explaining the architecture and design of the community version. This enabled the R&D team at General Data Technology to quickly grasp openGauss's architecture and design, and understand its differentiator advantages compared to foreign open-source databases.

For example, compared with the 32-bit xid capability of foreign open-source databases, openGauss provides a 64-bit xid feature. Since each database transaction needs to be assigned an xid, and as the transaction volume continues to increase, xid exhaustion can cause database freezing, thereby affecting normal business operations. “GBase 8c is a distributed database that handles transaction volumes several times or even dozens of times greater than centralized databases, making xid exhaustion much more likely. By changing xid from 32-bit to 64-bit, openGauss greatly reduces the risk of xid exhaustion, providing strong support for stable business operations,” said Zhang Yi, General Manager of Products at General Data Technology Co., Ltd.

During the system design phase, openGauss community technical experts reviewed the architecture design of GBase 8c, ensuring its design rationality and product competitiveness. In the development phase, the openGauss community provided one-on-one guidance to General Data Technology on specific code issues, improving development efficiency.

The current multi-model and multi-modal native distributed GBase 8c built on openGauss 3.0 delivers high performance, high availability, elastic scalability, and high security. It can be deployed on physical machines, virtual machines, containers, private clouds, and public clouds, providing secure, stable, and reliable data storage and management services for core systems in critical industries, internet business systems, and government and enterprise systems.

GBase 8c has been successfully deployed in projects such as health code systems, city commercial bank online lending platforms, identity authentication platforms, and market regulation commissions. For a government health code system in southern China, a new database was built with GBase 8c to solve the problem of displaying and scanning codes during peak hours, addressing performance bottlenecks at a low cost. In the financial and telecommunications sectors, GBase 8c has also completed POC testing and will progressively replace traditional centralized databases to support core business systems.

Looking ahead, on the technology front, GBASE (General Data Technology) will continue to invest R&D resources to jointly build the open-source database root community, making openGauss a mainstream open-source community in the database industry. On the market front, General Data Technology will collaborate with openGauss to launch more industry solutions, truly applying the capabilities of native distributed databases to the digital innovation and development of industries and enterprises, thereby contributing to the advancement of the database industry.