• Database: MySQL, PG, MongoDB.
• Migrated Objects:
  • MySQL/PostgreSQL: database(mysql), schema(pg), table, comment, index, sequence(pg), constraints.
  • MongoDB: collection, index, shardkey.

Example: MySQL -> MySQL / Mongo -> Mongo

Refer to MySQL task templates, MySQL tutorial, Mongo task templates, and Mongo tutorial.

Note

Structure migration is executed serially in a single thread. Notice the following configurations:

[extractor]
extract_type=struct

[sinker]
sink_type=struct
batch_size=1

[parallelizer]
parallel_type=serial
parallel_size=1

Failure strategy: interrupt(default), ignore.

• interrupt: If a particular migration fails, the entire task will be terminated immediately.

• ignore: If a migration fails, it will not affect the migration of other schemas, and the process will continue. However, the failure will be logged as an error.

[sinker]
conflict_policy=interrupt

Phased migration

In a complete data migration process that includes both structure migration and data migration, the task will be divided into three stages in order to accelerate data migration:

1. Migrate table structures + primary/unique keys ( necessities for data migration);
2. Data migration;
3. Migrate indexes + constraints.

Thus, we offer 2 types of filtering:

Migrate table structures + primary/unique keys

[filter]
do_structures=database,table

Migrate indexes and constraints

[filter]
do_structures=constraint,index

MongoDB structure types

[filter]
do_structures=collection,shardkey

MongoDB does not use a separate database structure type. collection creates selected collections and copies collection options, shardkey copies source sharding definitions. For sharded targets, connect through mongos; DTS runs enableSharding before shardCollection when needed. If the target is not mongos, shard key statements are ignored.