Data Vault-based Data Warehouse — Testing

All source data (even if "dirty") must be stored in the Raw Vault.

Validate that business rules applied in the Business Vault match the requirements. Testing Data Vault-Based Data Warehouse

Query a PIT table for a specific date in the past. Verify it correctly identifies the "active" Satellite records for that timestamp. 5. Business Vault & Logic Testing All source data (even if "dirty") must be

Insert a record with a modified attribute. Verify that a new Satellite record is created with the updated data while the old record remains (historical tracking). Verify that a new Satellite record is created

Ensure "Zero Keys" or "Ghost Records" exist in Hubs to handle late-arriving data or missing lookups without breaking the model. 3. Data Integrity & Reconciliation This ensures that "what went in is what came out."

Confirm Links correctly map relationships between Hubs. Test for referential integrity (though often not enforced by DB constraints in DV, it must be validated via query).

Since Data Vault is pattern-based, the first layer of testing ensures the "plumbing" follows the methodology.