OceanBase Migration Service (OMS) is a service that supports data interaction between OceanBase Database and homogeneous or heterogeneous data sources. It supports online migration of existing data and real-time synchronization of incremental data.
OMS provides a visualized and centralized management platform. You can migrate data in real time with simple configurations. OMS aims to help you achieve real-time data migration and synchronization from homogeneous or heterogeneous databases to OceanBase Database at a low cost and low risk.
Features and benefits
Support for multiple sources
OMS supports data transfer between OceanBase Database and multiple data sources, such as MySQL and Kafka. The features that OMS provides can be different depending on the types of source and destination terminals.
Real-time data migration without interrupting or affecting applications
You can use OMS to seamlessly migrate data to OceanBase Database without interrupting your application. After the application is migrated to OceanBase Database, all the changed data in OceanBase Database is synchronized to the original source database in real time.
OMS reduces the risks of business migration and helps enterprises build data architectures with high availability and reliability.
High-performance, secure, and reliable data migration
OMS replicates large amounts of data between heterogeneous IT infrastructures in real time with second-level latency. Therefore, OMS can be applied to multiple scenarios, such as data migration, inter-city data backup for disaster recovery, emergency systems, real-time data synchronization, disaster recovery, and database upgrade and migration.
OMS allows you to migrate and synchronize data without affecting or interrupting your business applications. It also ensures data integrity and transaction consistency. OMS supports full migration at a speed of 100 MB/s and delivers a transaction processing performance of 200,000 transactions per second (TPS), and a data synchronization performance of 50,000 requests per second (RPS). OMS provides a high-availability deployment architecture as well as stable and reliable transmission channels for data migration and real-time synchronization.
All-in-one interaction
OMS allows you to manage data migration throughout its whole lifecycle. You can create, configure, and monitor data migration tasks on the visual management page.
Real-time data synchronization to facilitate business decoupling
OMS supports real-time data synchronization between OceanBase Database tenants and self-managed Kafka and RocketMQ instances. It can be applied to business scenarios such as real-time data warehouse construction, data query, and report distribution.
Multiple data consistency checks
OMS provides multiple data consistency checks to ensure data quality. It displays differences in data to help you fix errors.
Convenient data migration to or from OceanBase Database
OMS supports the features of every version of OceanBase Database. It collects and converts database objects and supports data migration without interrupting the services of the source database.
Scenarios
This section describes the enterprise-class data migration and synchronization scenarios of OceanBase Migration Service (OMS).
In traditional database migration solutions, data is usually migrated during downtime to ensure the migration stability and data consistency. During the downtime, data writes to the source database must be temporarily stopped. The business database can be switched only after the migration is completed and the data consistency is verified.
The downtime duration varies according to the business data volume and network conditions. In case of a massive volume of business data, the downtime may last several days, greatly affecting the business.
OMS supports in-service data migration, ensuring service continuity of the source database during the migration and minimizing the impact of data migration on the business. During schema migration, full data migration, and incremental data migration, the full data and incremental data in the source database are synchronized to the sink database in real time. After the data verification succeeds, the business can be handed over from the source database to the sink database.
Real-time data synchronization
OMS can synchronize incremental data from OceanBase Database and other databases to message queues of self-managed instances such as Kafka and RocketMQ instances in real time. This improves the message processing capability and extends the business application in monitoring data aggregation, streaming data processing, online/offline analysis, and other big data fields.
OMS enables real-time data synchronization between physical tables of OceanBase Database and data sources such as Kafka. You can use OMS in cloud BI, real-time data warehouse construction, data querying, report distribution, and other business scenarios.
High availability and flexible deployment
The Store, Connector, and JDBCWriter components support high availability. When the Store, Connector, or JDBCWriter component fails, the corresponding abnormal process is restarted.
IDC-level disaster recovery is supported.
When a server fails, all components on this server are migrated to a functioning server.
When resource metrics on a server are abnormal, some components are migrated to a functional server based on the load balancing strategy.
When a store fails but the resources on the server are sufficient, OMS tries to restart it. If the resources on a server are insufficient, OMS implements load balancing or deletes redundant stores.
When a connector or JDBCWriter fails, OMS restarts the abnormal process.
Regional IDC-level disaster recovery is supported.
OMS supports separate deployment of incremental log read and incremental synchronous write. By using optimization methods such as multiple Transmission Control Protocols (TCPs) for parallel transmission, data compression, and data encryption, OMS can overcome challenges such as long distances, poor network conditions, and data security. Note that cross-region scheduling is not supported currently.
Features
OMS supports data migration and synchronization.
Data migration: A data migration project is a one-time task. After a data migration project is completed, resources occupied by the project are released. You can create a data migration project to migrate data between homogeneous or heterogeneous data sources in business scenarios such as database upgrade, cross-instance data migration, database splitting, and database scaling.
Data migration projects are the basic unit of the data migration feature of OMS. When you create a migration project, you can specify the migration level, which ranges from table to database. For more information about the project types and migration types supported in data migration, see Data migration.
Data synchronization: Data synchronization is a continuous operation. After a data synchronization project is created, data will be synchronized continuously to ensure consistency between the source and destination databases as well as the real-time flow of data of key business systems. You can create a data synchronization project to synchronize data between data sources in real time in business scenarios, such as multi-site high availability, remote disaster recovery, data aggregation, and real-time data warehousing.