Scheduled Scan

Reverse ETL (Scheduled Scan) continuously pulls data from the source and writes it to the target on a fixed schedule. This approach is ideal for:

• Data sources that lack CDC (Change Data Capture) logging capabilities.
• Scenarios requiring data refreshes at specific intervals (like every N seconds or minutes).
• Near real-time synchronization where slight delays are acceptable.

Supported Pipelines

Source -> Target	Sync Mode	Pre-execution Action (Optional)	Incremental Field1 Type
StarRocks -> MySQL	Scheduled Scan	Truncate Target2, Rename and Create3	DateTime
Doris -> MySQL	Scheduled Scan	Truncate Target, Rename and Create	DateTime, High-Precision DateTime
Redshift -> MySQL	Scheduled Scan	Truncate Target	Timestamp
Elasticsearch -> Elasticsearch	Scheduled Scan	Truncate Target	Date, Nanosecond Date, ES Timestamp
OssFile -> PostgreSQL	Scheduled Scan	Truncate Target4	/
SshFile -> PostgreSQL	Scheduled Scan	Truncate Target	/
S3File -> PostgreSQL	Scheduled Scan	Truncate Target	/
Yuque -> PostgreSQL	Scheduled Scan	Truncate Target	/

info

1. The Incremental Field requires a time-type column. Once configured, BladePipe automatically extracts newly added or updated records where the time is "after the last scan". This achieves incremental synchronization. (Note: This method cannot sync hard-deleted records from the source.)
2. Truncate Target: BladePipe empties the target table before starting the current scan. This is useful for periodic full table migrations (such as report results).
3. Rename and Create: BladePipe executes rename table A to A_arc_timestamp and create table A like A_arc_timestamp before starting the current scan. This is also useful for periodic full table migrations.
4. For file-based sources, "Truncate Target" only deletes the corresponding file data rather than emptying the entire table.

Operations

Select Data Sources

1. Select the source and target data sources, and click Test Connection.
2. Select the specific schema you want to synchronize.

Configure DataJob

1. On the DataJob page, select Incremental as the DataJob Type.
2. Select Periodic Scan as the Sync Mode, and set the parameters:
   • Interval: The interval between each scan in seconds (defaults to 300 seconds).
   • Pre-execution Action: The action you want to perform before starting a scan.

Select Data Tables

1. On the Tables page, select the tables you want to migrate.
2. Click Open Action Blacklist to filter out specific DML/DDL operations as needed.

Process Data

1. On the Data Processing page, select the columns you want to migrate.
2. Configure the Incremental Field as needed. If you skip this, BladePipe performs a full table scan every time.
   • Set individually: Click Operation > Incremental Field to set it for a single table.
   • Set in batches: Click Batch Operation > Incremental Field to apply the setting to multiple tables.

Confirm Creation

1. On the Creation page, verify all DataJob information.
2. Click Create DataJob.

Recommended Settings

• Interval: Start with a larger interval (such as 300 seconds). You can decrease it later once the pipeline runs stably.
• Pre-execution Action:
  • If you want to perform an incremental sync (by setting an Incremental Field on the Data Processing page), select None.
  • If the tables are small and you need to replicate source deletions, skip the Incremental Field and select Truncate Target instead.
• Incremental Field: Prioritize time fields that are sortable, clearly incremental, and comparable.

FAQs

Q: Why can't I see the "Incremental Field" option?

Your current data pipeline does not support this feature, or the selected field type is not supported.

Q: Can I skip configuring an Incremental Field?

Yes, if your data pipeline supports it. The system will generally perform a full scan or replication on every cycle.

Q: How do I know if the DataJob is running in RETL mode?

A distinct "clock" icon (Scheduled Scan) appears in the DataJob list and DataJob details page.