Pre-check is a necessary stage before the start of a migration task. It involves checking pre-conditions that may affect a successful migration, such as the connectivity of the source and target instances, and permissions of migration accounts. If a pre-check fails, you can troubleshoot the issues, and re-run the pre-check.
Schema migration indicates the migration of structure object definition syntax, including tables, views, triggers, stored procedures, storage functions, synonyms, and other structure object syntax. For migration between heterogeneous databases, data types are mapped in the schema migration stage, and object definition syntax is adjusted according to the source and target instance definition syntax. For example, in Oracle to MySQL migration, the number type in Oracle is mapped to the decimal type in MySQL.
Existing data migration indicates migration of all the existing data in the source instance database (except for the definition of structure syntax) to the target instance. If you only select existing data migration when creating a migration task, then during the migration process the new data written to the source instance is not migrated to the target instance.
Incremental data replication indicates synchronizing the incremental data written to the source instance during the migration process to the target instance. If you select existing data migration and incremental data replication when creating a migration task, DTS first implements static snapshot on the source instance, migrates the snapshot data to the target instance, and then synchronizes the incremental data written to the source instance during the migration process to the target instance. Incremental data replication is a process of real-time data synchronization between the source instance and target instance. The task cannot stop automatically. If you want to stop the migration, you must manually stop the task in the console.
Synchronization initialization indicates initializing the historical data of the object to be synchronized to the target instance before incremental data synchronization.
The synchronization initialization types can be divided into structural initialization and full-data initialization. Structural initialization is the initialization of the structure definition of the object to be synchronized. Full-data initialization is the initialization of the data of the object to be synchronized.
Synchronization performance indicates the number of records synchronized to the target instance per second, and the unit is RPS. Synchronization performance is a specification used for data synchronization service. Different specifications of performance have different number or records synchronized per second. For more informatioin, see Synchronization specifications.
Synchronization latency indicates the time difference between the timestamp of the latest data synchronized to the target instance in the source database, and the current timestamp in the source instance. Synchronization latency reflects the data time difference between the target and source instances. When the synchronization latency is zero, it indicates the source instance and target instance are fully consistent.
Subscription channel ID is the unique identity of the subscription channel. After you purchase a subscription channel, DTS automatically generates a subscription channel ID. When using SDK to consume incremental data, you must provide the corresponding subscription channel ID. The subscription channel ID corresponding to each subscription channel is displayed in the subscription list.
DTS divides data in the database into two types: data update and structure update. Data update modifies the data, but does not modify the structure object definition, such as insert, update, and delete.
DTS divides data in the database into two types: data update and structure update. Structure update indicates the modification of the syntax defined by the structure object, such as create table, alter table, and drop view. You can select whether to subscribe to the structure update when configuring a subscription channel.
Data scope is the range of incremental data timestamp stored in the subscription channel. The incremental data timestamp indicates the timestamp when the incremental data is applied and written to the transaction log in the RDS instance. By default, only incremental data of the most recent day is retained in the subscription channel. DTS regularly cleans expired incremental data and updates the data scope of the subscription channel.
The consumption time point is the timestamp corresponding to the latest consumed incremental data subscribed to downstream SDKs. The SDK reports the ACK to DTS for every piece of data it consumed. The server then updates and saves the consumption time corresponding to the SDK. If the SDK has an exception and restarts, the server automatically pushes the subscription data from the last consumption point.