Forum Discussion
Regarding Data Load issue
- 2 months ago
Hi Thirumala,
Thanks for reaching out.
"The resultset…has exceeded the maximum allowed size of '1,000,000' rows" is the 1-million-row intermediate limit.Key point: it's not the 15 measures. Aggregations like SUM/COUNT get pushed to the source and return small results. The error means something in the visual makes Power BI pull a pre-aggregation result set of over 1M rows — usually:
- A high-cardinality field on rows/columns (an ID, order number, timestamp) being grouped.
- A measure that needs all distinct values of a high-cardinality / degenerate-dimension column (stored at line grain).
- Measure or Top N filters, which fetch all categories from the source before filtering.
the fix is to reduce the rows returned:
- Find the culprit first. Performance Analyzer → copy the visual's DAX → paste into DAX Studio and check the generated SQL. It shows exactly which column forces the large result. Don't optimise blind.
- Remove or raise the grain of that high-cardinality field (group by month/category, not line-level).
- Filter early so the intermediate set stays under 1M.
- Aggregation table (Import mode) holding just that column's distinct values, if a measure genuinely needs it — facts stay DirectQuery.
References:
https://learn.microsoft.com/en-us/power-bi/connect-data/desktop-use-directquery
https://learn.microsoft.com/en-us/power-bi/guidance/directquery-model-guidanceBest regards,
Parchitect · Solutions Architect · Microsoft Fabric Specialist
Hi thiru2010
Direct Query has a hard limit of 1M rows per query, so a matrix with 40 tables and 15 measures is too heavy.
Best practices are reducing dimensions in the matrix. Aggregating data in the source. Using Import or Composite models for large visuals. Splitting measures across multiple visuals. Applying slicers and filters to cut the row counts.
For your reference:
https://community.fabric.microsoft.com/t5/Service/Direct-Query-1-million-row-limitation-in-Power-BI-Premium/m-p/847202