Troubleshooting

Troubleshooting: my Pipeline procedure will not generate

These problems happen on the destination schema’s Pipeline tab. For the normal flow, see Generate the load procedures.

“Procedure generation is not supported for dialect …”

Cause. The destination is PostgreSQL, MySQL, Snowflake, or Redshift. Load generation works only for SQL Server, Fabric Warehouse, Databricks, and Fabric Lakehouse.

Fix. If one of the supported targets works for you, change the Database System with Re-run, then finalize and generate again. Otherwise the destination schema is still usable for DDL; write the loads yourself.

“… requires a watermark column” / “… requires a source batch column” / “… requires a change detection method”

Cause. The strategy’s configuration is incomplete. The row shows a warning icon next to the strategy.

Fix. Click the warning icon or the gear, fill in the required field, click Save, then generate again.

A table is skipped, or its preview button is disabled

Cause. The table’s strategy is None. Regenerate All reports it as skipped (no strategy).

Fix. Pick a strategy for the table.

“Column … is set to MaterializeViaView but has no expression”

Cause. A calculated column is produced through a helper view, but neither the destination column nor the mapping has an expression.

Fix. Give the column an expression with Override expression, or choose a different handling such as Skip in pipeline. See Override a mapping with an expression.

“No generated procedures exist for this promotion. Generate child procedures first.”

Cause. You clicked Generate Master Proc before generating any table loads. The button is normally disabled until one exists.

Fix. Click Regenerate All first.

“All N procedure generations failed: …”

Cause. Regenerate All failed for every table it tried. The message lists the errors. When only some tables fail, the dialog finishes with a count such as 10 generated, 2 error(s) instead.

Fix. Preview one table to see its specific error, fix it, and try again.

The status still says Stale

Kenseme does not regenerate on its own. Click Regenerate All, or the refresh button on the row. The N stale badge in the header counts what is left.

I can’t find the generated procedures

They are hidden by default. On the destination schema’s Procedures, Scripts, or Views tab, tick Show Generated.

The generated code references a column that no longer exists

Fix. In the mapping editor, click Sync from Source, resolve stale mappings, validate, finalize, and regenerate.

The generated SQL or notebook fails when I run it

See Troubleshooting generated SQL or Troubleshooting the PySpark notebook.