MetricSign
EN|NLRequest Access
High severityexecution

Power BI Refresh Error:
DELTA_LIVE_TABLES_FLOW_FAILED

What does this error mean?

An individual Delta Live Tables flow (a single dataset definition within a pipeline) failed during an update, preventing that table from being written.

Common causes

  • 1A SQL or Python transformation in the flow definition raised a runtime exception
  • 2A source table referenced in the flow was deleted or has an incompatible schema
  • 3An expectation with 'fail' action mode was violated by incoming data
  • 4A join or aggregation on large datasets caused an out-of-memory error

How to fix it

  1. 1Step 1: In the DLT pipeline event log, click the failed flow to see the exception message and stack trace.
  2. 2Step 2: If a schema mismatch, update the flow definition to handle the new source schema or add schema evolution settings.
  3. 3Step 3: If an expectation failure, review the violating rows in the quarantine table (if configured) and decide whether to relax the expectation or fix the data.
  4. 4Step 4: Deploy the corrected flow definition and trigger a new pipeline update.
  5. 5Step 5: Add unit tests to the flow logic to catch regressions before they reach production.

Frequently asked questions

Does a single flow failure stop the entire DLT pipeline?

It depends on configuration. If other flows depend on the failed one, they also fail. Independent flows may still succeed. Terminal pipeline failures occur when a critical flow fails.

Where are quarantined rows stored when a DLT expectation fails?

You must explicitly configure a quarantine table in your DLT flow using EXCEPT ROWS WHEN... syntax. If not configured, violating rows are simply dropped or the flow fails depending on the expectation action.

Other execution errors