All Products
Document Center

Data migration

Last Updated: Dec 20, 2017

Data migration helps you quickly and easily migrate data between different databases. It is applicable to scenarios such as on-premises data to cloud migration, Alibaba Cloud cross-instance migration, and database split and expansion. Both homogeneous migration, such as MySQL to MySQL, and heterogeneous migration, such as Oracle to Alibaba Cloud RDS MySQL, are attainable. DTS also provides basic ETL functions, including multi-level object mapping and data filtering.

Migration types

The complete data migration process consists of three sequential stages. When creating a migration task, you can determine which stages are most applicable to your environment. These stages are:

  1. Schema migration: Migrates structural definition from the source instance to the target instance.

  2. Existing data migration: Migrates the existing data from the source instance to the target instance.

  3. Incremental replication: Replicates the source instance data changes to the target instance. The data changes are captured from the start of the existing data migration until the end of the whole migration progress.

Note: All these stages of data migration are required if you want to perform zero downtime data migration.

Migration task

Migration task is the basic unit of data migration. To perform data migration, you must create a migration task in the DTS console. With a migration task, you must specify the connection between the source and target instances, the objects to be migrated, migration stages, and additional parameters. You can also create, manage, suspend, or delete a migration task in the DTS console.

The migration task displays different status during creation and progress.

Migration status Description Available operation
Not started The migration task has been configured, but is not started. Pre-check
Pre-checking The migration task is performing pre-check. Delete
Pre-check passes The migration task passes pre-check. Start
Migrating The data is being migrated. Pause
Migration Failed The migration failed. Delete
Paused The migration is suspended. Start
Finished The migration is complete or cancelled. Delete

Migration specification

Data migration supports the following list of migration specifications, in form of source database engine to target database engine.

Migration Specification Schema Migration Existing Data Migration Incremental Data Replication
Oracle > RDS For MySQL Support Support Not support
MySQL > RDS For MySQL Support Support Support
SQLServer > RDS For SQLServer Support Support Support
PostgreSQL > RDS For PostgreSQL Support Support Support
Redis > Redis Support Support Support

Source instance

You can use the following databases as the source instance for data migration.

  • RDS instance
  • On-premises databases
  • ECS built-in database

Target instance

You can use the following databases as the target instance for data migration.

  • RDS instance
  • ECS built-in database
  • Redis instance

Multiple ETL functions

Data migration supports multiple ETL functions, including:

  • Multi-level object mapping: You can perform data migration between source and target objects with different names by configuring mapping relationship in the DTS console. This feature supports objects such as databases, tables, and columns.

  • Data filtering: You can filter the table to be migrated by configuring a standard SQL condition. For example, by specifying the time period, you can only migrate the latest data.

DTS monitor

DTS monitors the operating status of migration tasks and detects abnormalities, such as lag or abended processes. When an abnormality is detected during the data migration process, DTS immediately sends an alarm message to you instructing you how to best handle the issue.