- PostgreSQL now supports the Change Data Capture (CDC) job type. CDC from PostgreSQL uses the logical replication slot configured in you source database to read Write-Ahead-Logs. More information on how to configure Logical Replication can be found here.
- Users can now select which Jobs to export when migrating. During migration, Sync will export:
- The required Connections, Tasks, and Events associated with each Job.
- The last run status for each Job and Task.
- Added the ability to archive logfiles instead of deleting them. Logs can be archived to another folder on disk or uploaded to Amazon S3.
- The command timeout is now configurable for Transformations. The default value is 300 seconds.
- The embedded Jetty WebServer now supports JNDI Connections.
- The CACHE Statement now supports standard Replicate Options in the Query. For example:
CACHE [DestinationTable] WITH AutoTruncateStrings='True' SELECT * FROM [Table]
- Google Cloud destination connections have been migrated to MySQL or PostgreSQL connection types. Existing connections will be unaffected.
- All Connection types that were previously only in the Professional or Enterprise tier have been moved into the Standard tier. The current connection types in the Standard and Starter tiers are unchanged.
- Add Tables now lets you choose a schema. This only applies to source connection types that support multiple schemas, such as SQL Server and PostgreSQL.
- When the source table includes a schema, Add Tables will build out the full replicate command with the Select statement. This change should mitigate some confusion when both source and destination schemas are being used.
- Empty logfile column values now contain the value N/A. This change mostly affects transformations that use connectors that cannot generate logs.
Source and Destination Connectors
- Added support for the UUID datatype in Azure Synapse.
- Added support for the BigNumeric datatype in Google BigQuery.
- Added support for CustomSchemas for NetSuite connections.
- Starting with Hubspot connector version 21.8076, some Hubspot primary keys changed from type int to type long to avoid overflow errors. Tables created with a previous version of the connector will need to be recreated with the new data types. To do this, run the Job or Tasks with Drop Table enabled.
- Added the WaitForResults parameter to the ExecuteJob endpoint. This parameter controls whether Sync will wait for the Job to complete and return the results of the run.
- Deprecated the Async parameter from the ExecuteJob endpoint.