Data Transfer Service (DTS) provides database data transmission as a bundle, including data migration, data sync, and data subscription. It helps you easily complete database migration without interrupting business services and use real-time synchronization channels to build a high-availability database architecture for remote disaster recovery.
The key APIs used in a data migration task are as follows. The circles represent the migration task in different statuses. Perform the API invocation operation sequentially according to the serial numbers in the figure.

In a migration task, the supported operations and relevant APIs for each task status are as listed in the table below:
| Task Status | Supported Operations | Key API | Description |
|---|---|---|---|
| - | - | Purchase Migration Service CreateMigrationService | Purchase migration task resources. After purchase completion, a randomly generated migration task ID list will be returned. |
| Creation completed | Configuration | Configure migration service ModifyMigrationJob Validate migration task CreateMigrateCheckJob Query data migration task list DescribeMigrationJobs |
Purchase operation completed. Next, configure the migration task. After task configuration, initiate verification. Task verification validates parameters of the source and target databases before migration task startup. The migration task can start only after verification passes. Use "Query data migration task list DescribeMigrationJobs" to query all migration tasks and obtain the task ID. |
| checking | - | Query migration verification task results DescribeMigrationCheckJob | The migration task check is in progress. |
| Verification passed | Start now, Configure, Verify | Start data migration task StartMigrateJob | The migration task check passed, start the data migration task. |
| checkNotPass | Configuration, Verification | Configure Migration Service ModifyMigrationJob | The migration task check failed. You need to reconfigure the migration task and then initiate verification again. |
| Preparing readyRun | Termination | - | Transitional state after performing a startup operation, preparing to start up the data migration task. |
| Running | Termination, Create Consistency Check Task (Only supported for operable links with this feature) | Terminate Data Migration Task StopMigrateJob | The task is running. If the user does not need to continue the migration, the task can be canceled through the terminate operation. The task cannot be restored after termination. |
| stopping (terminating) | - | - | Transitional state after performing the termination operation. |
| canceled | - | Isolated Data Migration Service IsolateMigrateJob | canceled tasks need to be isolated so that the unfreezing fee can be processed promptly (pay-as-you-go billing will freeze the hourly fee). |
| Ready Complete readyComplete | Complete, Terminate, Create Consistency Check Task (Only supported for operable links with this feature) | Complete Data Migration Task CompleteMigrateJob Create Consistency Check Task CreateCompareTask |
For the "Full + Incremental Migration" scenario, when the task status is "Ready Complete", it indicates that both full and incremental data migration is completed. Users need to select a suitable time to end the task via the Complete operation. It is recommended to verify data consistency before the Complete operation. Data consistency verification compares the data selected for migration from the source database with the data result migrated to the target database to check for consistency and view details of any inconsistencies. Upon successful creation of a consistency check task, the task ID for the consistency check will be returned. |
| completing | - | - | Transitional state after you perform the complete operation. |
| Task success | - | Isolated Data Migration Service IsolateMigrateJob | The migration task completed smoothly. Completed tasks must be isolated so that the unfreezing fee can be processed promptly (pay-as-you-go billing freezes the hourly rate for one hour). |
| Task error | Stop, Retry | Retry migration task ResumeMigrateJob | During the migration process, when a task encounters an error and is interrupted, and the task status is "error", the task retry process can be used. |
| Retrying resuming | - | - | Transitional state after performing a retry operation. If retry successful, task proceeds normally; if retry failed, task status is "Task failure failed". |
| Task failure | Retry (Only Redis migration is retryable) | Isolated Data Migration Service IsolateMigrateJob | Migration task failure. failed tasks need to perform isolation operation so that the unfreezing fee can be processed promptly (pay-as-you-go billing will freeze the hourly fee). |
| All States | View, Resize, Isolate | Query a Data Migration Task Details DescribeMigrationDetail Adjust Instance Specification ModifyMigrateJobSpec Isolate Data Migration Service IsolateMigrateJob |
Use the View operation to view details of a designated migration task. Use the Resize operation to adjust the specifications of a migration link; transmission rates vary for different specifications. Use the Isolate operation to isolate completed tasks, failed tasks, and unnecessary tasks. |
| Isolated (billing status) | De-isolate (Restore), Decommission | RecoverMigrateJob DestroyMigrateJob |
A task will enter isolated state if a user-initiated isolation operation is performed or if it is in arrears for 1 day. Resources in isolation will be retained for 7 days, during which business downtime occurs and they are unavailable. After 7 days, the resources will be purged. For an isolated task, users can proactively initiate a decommission operation, or wait for the system to automatically take it offline after 7 days of isolation. If a user misoperation causes isolation, the task can be restored to its pre-isolation state through a de-isolate operation. |
The key APIs for use in data sync tasks are as follows. The circle represents the sync task in different statuses. According to the serial number in figure, perform the API invocation operation sequentially.

In a sync task, the supported operations and relevant APIs for each task status are as listed in the table below:
| Task Status | Supported Operations | Key API | Description |
|---|---|---|---|
| - | - | CreateSyncJob | Purchase synchronization service resources. After purchase completion, a randomly generated Task ID list will be returned. |
| UnInitialized | Configuration | Configure Synchronization Task Create Verification Synchronization Task Query Synchronization Task Information |
The purchase operation is complete. Next, configure the synchronization task. After the task configuration is completed, initiate verification. Task verification involves checking the parameters of the source and target databases before starting the synchronization task. The synchronization task can only be started after the verification passes. You can query all synchronization task lists and obtain the task ID via "Query Synchronization Task Information". |
| Checking | - | DescribeCheckSyncJobResult | The sync task verification is in progress. |
| Verification Passed | Start, Configure, Validate | Start Sync Task | The sync task verification passed, and the data sync task started. |
| Verification Failed | Configuration, Verification | Configure Synchronization Task | The synchronization task validation failed. You must reconfigure the synchronization task and then initiate verification again. |
| Preparing | End | - | Transitional state after startup operation, preparing to start data synchronization task. |
| Running | End | StopSyncJob | The task is running. If the user does not need to continue synchronization, they can stop the task through the End operation. |
| Stopping | - | - | Transitional state after performing the stop operation. |
| Stopped | - | Isolated Synchronization Task | The sync task completed smoothly. Completed tasks need to be isolated so that the unfreezing fee can be processed promptly (pay-as-you-go billing freezes the hourly fee for one hour). |
| Task error ResumableErr | End, Retry | Retry synchronization task ResumeSyncJob | When a task encounters an error during the sync process, causing task interruption and the task status is "ResumableErr", the task can be retried. |
| Retrying Resuming | - | - | Transitional state after performing a retry operation. If the retry is successful, the task resumes normal operation; if the retry fails, the task status is "task failure Failed". |
| Task Failure | - | IsolateSyncJob | The synchronization task failed. Failed tasks need to undergo an isolation operation so that the unfreezing fee can be processed in a timely manner (pay-as-you-go billing will freeze the hourly fee). |
| All states | View, Adjust, Isolate | Query synchronization task information DescribeSyncJobs Adjust synchronization task specification ResizeSyncJob Isolate synchronization task IsolateSyncJob |
Use the view operation to view details of a certain assigned synchronization task. Use the adjust operation to adjust the sync link specification. Transmission rates vary for different specifications. Use the isolate operation to isolate completed tasks, failed tasks, and unnecessary tasks. |
| Isolated (Billing Status) | De-isolate (Restore), Decommission | RecoverSyncJob DestroySyncJob |
A task enters the isolated state either through user-initiated isolation operation or after being in arrears for one day. Resources in isolation are retained for 7 days, during which the service is unavailable. After 7 days, the resources will be purged. For an isolated task, users can proactively initiate a decommission operation or wait for the system to automatically decommission it after 7 days of isolation. If a user misoperates and isolates a task, they can perform a de-isolate operation to restore the task to its pre-isolation state. |
The key APIs used in data subscription tasks are as follows. The circles represent the subscription task in different statuses. According to the serial numbers in the figure, perform API calls sequentially.

| Task status | Supported operations | Key API | Description |
|---|---|---|---|
| - | - | Create Subscription Task CreateSubscribe | Purchase subscription task resources. After purchase completion, a list of randomly generated subscription task IDs will be returned. |
| notStarted | Configure Subscription, Destruction/Return Products | Configure Subscription Task ConfigureSubscribe Check Subscription Task CreateSubscribeCheckJob |
Purchase completed. Next, perform subscription task configuration. After task configuration is completed, initiate verification. The verification task checks various parameters of the source database before subscription task startup. The subscription task can start only after the verification passes. |
| checking | Query subscription verification results DescribeSubscribeCheckJob | The subscription task verification is in progress. | |
| Verification passed | Start up, configure subscription, validate, terminate/return products | Start validation task | Verification passed. Start the data subscription task. |
| Validation Failed | Configure Subscription, Verification, Termination/Return Products | Configure Subscription Task | Subscription task validation failed. Reconfigure the subscription task, then initiate verification again. |
| Starting | Terminate/Return products | - | Task starting, transitional state. |
| Task running | Modify Subscription Object, Reset Subscription, Destroy/Return, View Task Details | Create Consumer Group Modify Subscription Objects Reset Subscription Destroy/Return Subscription Task View Subscription Task Details |
When a subscription task is running, you can create a consumer group to consume Kafka. Modify Subscription Objects (ModifySubscribeObjects) allows you to add or delete objects, such as Kafka partition rules, for an ongoing task. Destroying a task stops it, clears the configuration and Kafka data, and performs billing settlement. Resetting a subscription task resets a task that is no longer needed. The old task stops, its historical configuration and Kafka data are cleared, and a new task is reconfigured. Resetting a task helps users avoid frequent purchase operations, such as refunding the old task and repurchasing a new one. While a task is running, you can view its details and operate relevant Kafka APIs. |
| stopping | - | When resetting the subscription object, the subscription task is stopping. After resetting, you can reconfigure the task only when the task status changes to "notStarted". | |
| Task error | Restore, Terminate/Return products | Resume subscription task ResumeSubscribe | During the subscription process, if a task encounters an exception causing task interruption and its status is "error", you can try to resume the task. |
| Isolated | Offline | OfflineIsolatedSubscribe | After a destruction or return operation, the task enters an isolated state. Users can proactively initiate an offline operation for an isolated task, or wait for system automation to take it offline after 1 day of isolation. |
フィードバック