Dmitry, The AWS product is not limited in the way it does aggregation. Anything which works on stand-alone JRS, will work on AWS. There is a setting you can change to push down aggregation, you can see if this addresses your issue (might depend on DB type): http://community.jaspersoft.com/wiki/unexpected-results-andor-performance-issues-group-summaries-ad-hoc-v50-and-later Alternatively, you could explore a couple other options: - Create a derived "aggregate" table for the target data in his domain, which groups the values as desired and returns count(distinct ColumnName) - Create a custom entry point/SQL function in the semantic layer config BTW, the 200K limit can be changed in Ad Hoc Server Settings (Manage -> Server Settings -> Ad Hoc Settings -> "Ad Hoc Dataset Row Limit".