According our OLAP businesses, many have more than 15 dimensions. And in fact
we can't optimize cube because the dimension-group changes frequently.

SO! We are thinking about that maybe we can no build all 15 dimensions at
step cube building.
We can only build a part of them such as 11 dimensions. 
Then when queried, we calcutale the result set according to the results of
11 cuboids.

Does anyone have some ideas? 
Maybe we can discuss about it.

--
View this message in context: 
http://apache-kylin.74782.x6.nabble.com/No-build-all-dimensions-tp7263.html
Sent from the Apache Kylin mailing list archive at Nabble.com.

Reply via email to