Functionality Principle

Full Data Migration

UDTS has deployed services in many regions around the world. It will select the running location based on the source and destination location to ensure migration efficiency.

One-way Incremental Sync

UDTS simulates a Slave node and retrieves incremental data from the Binlog starting from the full data export point.

Two-way Incremental Sync

Similar to one-way incremental sync, it simulates a Slave node to get incremental data. Meanwhile, UDTS marks the written data. When a new Binlog Event occurs, it first checks if there is a mark. If there is a mark, it indicates looped data, so it’s discarded directly, otherwise, it adds a mark and writes it to the opposite end.

Technical Advantages

• Operates as a service, users do not need to consider physical resources
• Supports various types of migrations
• MySQL to MySQL supports migration at different dimensions, including table, database, multiple databases, and the entire database
• MySQL to MySQL supports ETL
• MySQL to MySQL supports one-way incremental sync as well as two-way incremental sync. It supports syncing via GTID and binlog file/pos. Supports resumable upload.
• Migrating MySQL from AWS can be done directly from the master database, no need to set up a slave.
• Supports migration through intranet, internet, and dedicated lines.
• Provides migration speed limit functionality.

Transmission Speed

Operational efficiency of a single UDTS task in different scenarios. UDTS can run concurrently and provides flexible migration methods, supporting migrations by database or by table. Migration rate is typically limited by migration bandwidth or the performance of the source and destination databases. By using multiple tasks in parallel, the migration bandwidth can generally be utilized 100% in most scenarios.

One-way incremental sync peak supports 20,000 OPS (insert, update, delete)