Troubleshooting

Troubleshooting: my Pipeline will not validate

Find the message from the Validation Results dialog below. For how validation works, see Validate a Pipeline.

“…load strategy … requires at least one key column”

Cause. The table’s strategy merges rows by key (Incremental Load, Incremental Batch, Incremental Hash, SCD Type 2), but no included column is marked as a key. The No PK badge on the table row shows the same thing.

Fix.

  1. In the mapping editor, expand the table.
  2. Tick Is PK on the column or columns that identify a row.
  3. Click Validate again.

If the table has no natural key and rebuilding it each run is acceptable, switch it to Truncate & Reload instead. Ticking SK does not satisfy this check.

“View … has no primary key designated”

Cause. Every source view becomes a destination table and needs a key.

Fix. Expand the view’s row and tick Is PK on a column. Kenseme adds a key column named <Table>Id to every view when the Pipeline is created; if you cleared it, tick it again.

“…requires a change detection method (Hash, Watermark, or Batch)”

Cause. An SCD Type 2 table has no change detection method.

Fix. On the destination schema’s Pipeline tab, click the warning icon or the gear on the row, choose a Change Detection Method, fill in its setting, and click Save.

“Duplicate destination table name …”

Cause. Two included tables map to the same destination name, often because two source schemas have a table with the same name.

Fix. Rename one of them in the Destination Table column, or exclude one.

“Table … has no destination name” or “Column … has no destination name”

Fix. Click the empty destination cell and type a name.

“Column … has no length for NVARCHAR — …”

Cause. SQL Server, Fabric Warehouse, MySQL, and Redshift need a length on text columns, and the source column has none.

Fix. Give the source column a length in the source schema, then click Sync from Source and validate again.

🚩 TODO: Kenseme can store a destination type override per column, but the mapping editor shows Destination Type read-only. Confirm whether users should be able to set a destination length in the Pipeline.

“Table/column …: … — run Sync from Source → Resolve stale mappings.”

Cause. The source table, view, or column no longer exists.

Fix. Click Sync from Source, then Resolve stale mappings in the banner. See Keep a Pipeline in sync with its source.

Type compatibility errors

Cause. A destination data type cannot hold the source type. Narrowing conversions that may lose data are warnings; impossible or unrecognized conversions are errors. The message names the table, column, and reason.

Fix. Align the types in the source or destination schema, or exclude the column. The Summary tab’s Data Type Conflicts section lists type changes across the Pipeline.

Warnings worth acting on

  • “watermark column not configured” or “batch column not configured”. Validation passes, but generating that table’s load will fail. Configure the strategy on the Pipeline tab.
  • “Source and destination schemas overlap — TruncateAndReload will destroy source data”. The destination is also a source. Pick a different destination or avoid Truncate & Reload on those tables.
  • “N AI suggestions have not been reviewed”. Accept or reject them, or ignore the warning.
  • “… has no primary key selected”. Allowed for Truncate & Reload, but consider adding a key.

Validation passes but Finalize is still disabled

Your organization may be read-only. See Account states and read-only mode.