Skip to main content
FULL_TABLE rereads the selected source stream for a run. Use it for:
  • first-run validation
  • controlled manual refreshes
  • debugging source and destination shape
  • confirming SQL model output before moving to Incremental

​
How to configure

  1. Open the Destination panel.
  2. In Config, choose FULL_TABLE.
  3. Keep write mode as Upsert.
  4. Save the destination.
  5. Run manually from the Destination node.

​
Scheduling

Full Table destinations are manual-only in the current app. The Scheduling tab is disabled until the destination uses INCREMENTAL.

​
Run behavior

Full Table still follows the strict delivery contract:
  • source rows are staged in temporary storage
  • SQL models run against raw staged tables
  • final destination tables must already exist
  • temporary staging is cleaned up after checkpoint extraction