What to check
A join can multiply rows when the participating data is not unique at the assumed key. The query may become slow, consume excessive memory, or produce unexpected totals. A cardinality label describes an assumption; changing the label does not make the underlying records unique.
Troubleshooting steps
- Identify the affected join and compare the grain and complete business keys on both sides, including time-dependent or language-dependent data.
- Use a small representative selection to check matching rows and duplicate keys before aggregation. Review the filters and join conditions.
- Correct the proven model or data-quality issue in development, documenting the intended relationship and validating any cardinality setting against real data.
Confirm the result
Compare row counts and key-figure totals before and after the change. Keep unmatched and duplicate cases in the test set so an apparently faster join does not silently discard business records.
References
The SAP KBA link provides a public symptom preview; its full resolution requires SAP Support access. These diagnostic steps are independent guidance. Check the installed release before applying a correction.