Background information

PolarDB provides a database solution for one of the top three insurance companies in China. The insurance company wants to migrate the major business and business systems to databases developed by Chinese vendors. This project is considered to be one of the largest projects for data migration to databases developed by Chinese vendors. The company builds a layered digital platform based on cloud computing and Internet-based technology. This platform uses a unified technical architecture, follows unified standards, and enables the interconnection between internal and external systems. This platform uses a powerful computing engine and is designed for lightweight applications. The company collaborates with IT innovation sectors to handle technical difficulties and takes the lead in building a hybrid cloud infrastructure. In the layered architecture, databases are the major components of the system. Databases are also the source of various technical difficulties for IT innovation. Therefore, the company plans to perform database transformation to make a breakthrough in IT innovation. The company plans to use PolarDB to replace all databases within one year. The duration of this project may change based on actual conditions. In this project, PolarDB O Edition supports high compatibility with the features of original production databases and provides a scientific and efficient database solution for the company. PolarDB helps the company make an overall business plan and steadily implement the plan.

Solution

The data center and the R&D center of the company work together with the Alibaba Cloud PolarDB O Edition team to formulate a detailed plan for smooth migration across business systems. This also reduces security risks during migration. For major business, the company applies a primary/secondary database architecture. Data Transmission Service (DTS) is used to write data back to the original secondary production databases in real time. After the new database systems properly run for a period of time, the company can replace the original production databases. This ensures seamless switchover. During this process, no databases are switched back to the original production databases. DTS is also used to synchronize incremental insurance data in PolarDB O Edition to Apache Kafka in real time. The change tracking feature is used to monitor the downstream business systems and the big data platforms to meet the requirements of diverse scenarios. Before the database systems are officially put into use, the technical team evaluates the compatibility in a comprehensive way, runs SQL regression tests during the rollout, and provides adequate monitoring and efficient emergency response methods after the rollout. This ensures smooth switchover between databases.

The company has migrated the major business to PolarDB O Edition and plans to deploy primary databases and secondary databases in the China (Beijing) and China (Shanghai) regions. Cross-region deployment of databases meets the requirements of geo-disaster recovery. Remote data replication between primary and secondary databases is used for cross-zone disaster recovery. This meets the requirements of the financial industry.

The database management system (DBMS) of PolarDB uses the multitenancy architecture to support multiple business lines. PolarDB allows you to centrally manage databases.

Advanced Database & Application Migration (ADAM) and DTS are used to migrate data online. This makes migration efficient and cost-effective.

Highlights

  • The company uses PolarDB Cluster Edition. A PolarDB cluster uses the multitenancy architecture to support multiple business lines. This way, the company can deploy PolarDB clusters based on business lines. This can meet the requirements of different business groups. As a result, hardware resources can be efficiently managed and utilized.
  • PolarDB uses a storage area network (SAN). This ensures high scalability, reduces O&M costs, and simplifies management.
  • PolarDB allows the company to enable high availability and disaster recovery in multiple modes. The supported disaster recovery modes include cross-zone disaster recovery and geo-disaster recovery.
  • PolarDB is highly compatible with the original production databases of the company. This way, the company can smoothly migrate data to PolarDB databases. Database users can use PolarDB databases in a similar manner in which they use the original production databases. The DBMS of PolarDB databases provides a simple method to connect to internal systems and monitoring and alerting platforms.
  • DTS can be used to write data back to the original production databases. This way, the company can switch business back to the original databases and synchronize data to stream-processing platforms to distribute data.

Benefits

This company is one of the largest insurance companies in China. The major business systems of this company have been migrated to self-developed databases in China. This suggests that the insurance company has taken an important step towards digital transformation based on independent technology. This proves that self-developed databases in China can meet the business requirements of the financial industry. From development to implementation, this project can be viewed as a typical case in the financial industry. The technology and components that are used in this project have reference and promotional value. After the data is migrated, PolarDB databases properly run for more than 150 days and no production incidents occur.

PolarDB provides the following major benefits for the company:
  • Compliance: The company can independently manage and control databases. PolarDB fully complies with the national requirements in China for database transformation in the IT innovation industry and helps the company avoid unexpected risks.
  • Cost: This project helps the company integrate multiple databases into PolarDB clusters. Hundreds of the original business databases of the company are integrated into PolarDB clusters. PolarDB allows the company to centrally manage databases. Data integration reduces the data volume by two-thirds. As a result, the company significantly saves hardware and O&M costs each year. To implement this project, the company does not need to significantly modify the original business systems. This saves costs for application rebuilding and simplifies database transformation.
  • High availability: PolarDB provides a zero recovery point objective (RPO) and reduces the recovery time objective (RTO) to less than 30 seconds. This delivers 99.999 percent data availability and reduces overall business risks.
  • Auto scaling: PolarDB automatically increases the number of resources during peak hours and reduces the number of resources after peak hours. This way, the resources can be shared across databases. PolarDB helps the company efficiently handle the surge in workloads during promotions.