Fivetran Release Notes

We have reverted the Khoros Communities connector to use the Bulk Data API v2 with HTTP Basic authentication. To continue syncing data, update your existing Khoros Communities connection with your Token, Client ID, and Community ID. For more information, see our setup instructions.

We have updated the sync strategy for the CAMPAIGN_STEP_STATISTIC table. We now incrementally sync the table to capture new records and updates. Previously, we re-imported the table to capture updates and deletes.

We have improved sync reliability by handling 404 responses when a webhook_id is no longer available. We now log a warning in the Fivetran dashboard, skip the affected API call, and continue the sync rather than fail it.

We now support both classic and multi-level boards in monday.com and sync all items and subitems across every hierarchy level.

We have improved sync reliability for the PRODUCT table and its child tables. In addition to incremental syncs, we now re-import these tables once a day to capture updates and deletes.

Our Oracle Fusion Cloud Applications connectors now support syncing CLOB data type columns. Previously, we skipped the columns with CLOB data types during sync. We now convert them to STRING data type in your destination tables. We are gradually rolling out these changes to all existing connections.

We've added the following new tables:

• COMPANY_INFO
• EXCHANGE_RATE

Our Appfigures connector can now fetch the historical data until January 1, 2009, for the following tables:

• AD_REPORT
• ADVERTISEMENT_SPEND
• ESTIMATE
• PAYMENT
• REVENUE
• SALE
• SUBSCRIPTION

Previously, we retrieved only the past 30 days of historical data for these tables.

We've fixed an issue where the REVIEW_CUSTOM_QUESTION_NAME table failed to sync. The table will now sync correctly. To backfill historical data for this table, trigger a re-sync on the REVIEW table.

We have added a new column, opens_proxy_excluded_unique_opens, to the CAMPAIGN_REPORT table.

We have fixed an issue where incremental syncs could skip records due to pagination limit. Syncs now use offset-based pagination to ensure all records are fetched reliably. Previously, syncs relied on a pre-computed row count to paginate results within each time window.

We have improved how we determine column types for Parquet files that our File connectors sync to your destination. We now read column types directly from the Parquet file's schema metadata. Previously, we inferred them from the actual values in the rows. As a result, you can now find column types that more accurately match the source in your destination schema.By default, this improvement is available to all new connections created on or after July 21, 2026. We are gradually rolling out this improvement to all existing connections.

We have added a new column, country_or_region_serving_state_reasons, to the CAMPAIGN_HISTORY table.We have added a new column, app_preview_device_with_assets, to the PRODUCT_PAGE_LOCALE table.

Our SAP OData connector is now generally available.For more information, see our SAP OData documentation and SAP OData Setup Guide.

We have changed our sync strategy for the PMSE_CASE table. We now incrementally sync this table instead of re-importing it.

You can now move connections from one destination group to another using the Move a Connection (Beta) REST API endpoint.

We have changed the sync strategy for the ACTIVITY_TYPE table. We now re-import this table once a day instead of incrementally syncing it.