What causes INTERNAL_ERROR (SQL Warehouse)?

Photon engine bug triggered by an edge-case query plan or data format. SQL warehouse pod crashed due to memory pressure from a very large result set. Transient cloud infrastructure fault during query execution. Known issue in a specific Databricks Runtime version — check release notes. Very large shuffle operation exhausting warehouse disk spill capacity

How do I fix INTERNAL_ERROR (SQL Warehouse)?

Retry the query — transient infrastructure faults often resolve automatically. Check the Databricks status page (status.databricks.com) for ongoing incidents. If reproducible, open a Databricks support ticket and include the query ID (visible in the query history). Try reducing the query complexity or adding explicit intermediate materializations to isolate the problematic plan. Upgrade or downgrade the SQL warehouse Databricks Runtime version to rule out a known engine bug. For memory-related INTERNAL_ERROR, increase the warehouse cluster size or enable result caching

High severityinfrastructure

Power BI Refresh Error:
INTERNAL_ERROR (SQL Warehouse)

What does this error mean?

The Databricks SQL warehouse (Photon or classic) encountered an unexpected internal error. Unlike FAILED in Jobs, this indicates a platform-level fault rather than user code.

Common causes

1Photon engine bug triggered by an edge-case query plan or data format
2SQL warehouse pod crashed due to memory pressure from a very large result set
3Transient cloud infrastructure fault during query execution
4Known issue in a specific Databricks Runtime version — check release notes
5Very large shuffle operation exhausting warehouse disk spill capacity

How to fix it

1Retry the query — transient infrastructure faults often resolve automatically
2Check the Databricks status page (status.databricks.com) for ongoing incidents
3If reproducible, open a Databricks support ticket and include the query ID (visible in the query history)
4Try reducing the query complexity or adding explicit intermediate materializations to isolate the problematic plan
5Upgrade or downgrade the SQL warehouse Databricks Runtime version to rule out a known engine bug
6For memory-related INTERNAL_ERROR, increase the warehouse cluster size or enable result caching

Frequently asked questions

Is INTERNAL_ERROR in a SQL warehouse the same as in Databricks Jobs?

Conceptually similar but contextually different. In Jobs it typically means a platform/cluster fault. In SQL warehouses it is almost always a Photon or warehouse-pod-level fault unrelated to user code.

Where do I find the query ID to include in a support ticket?

In the SQL Editor, click the query in history and copy the Query ID from the details panel. For programmatic execution, it is returned in the statement execution API response.