A query is efficient when involved dimensions all come from a single aggregation group. Or otherwise, large scale runtime aggregation will happen in memory, which is your error.
Try revise the aggregation group to fit your query pattern. Note - A dimension can appear in multiple groups. - Keep less than 10 dimensions in one group, or the cube will be huge. - Create a second cube with different aggregation group is also an option. We can help more if you can provide more details on your cube and queries. Cheers Yang On Wed, Apr 1, 2015 at 2:35 PM, Luke Han <[email protected]> wrote: > Forward to mailing list for further discussion. > > Thanks. > > 在 2015年4月1日星期三 UTC+8上午10:38:00，vipin kumar写道： >> >> Hi, >> >> I have limitation on aggregation group of 20 and so I created two >> aggregation group but when I am doing query b/w two different aggregation >> group its not working and also taking kylin down with heap space and out of >> memory error. >> >> Thanks >> Vipin Kumar > >
