Hi,
You can refer to http://kylin.apache.org/docs/install/configuration.html for some configs to speed up the step of dimensions building. Set "kylin.engine.mr.build-uhc-dict-in-additional-step=true" will trigger additional MR jobs to build dicts for UHC dimension, may this will reduce the time cost for building dicts in your env. For the question of dict reuse, you can set "kylin.dictionary.growing-enable=true" to enable it. Best regards, Ni Chunen / George On 09/9/2019 11:17，Guosheng Lu<[email protected]> wrote： https://mp.weixin.qq.com/s/DKMKTUV1xgrpbFau5L4baw this may be useful for resolve your question 在 2019/9/9 上午9:48，“wonder87”<[email protected]> 写入: I have mutiple cubes which contains some same dimension such as date dimension and user dimension , user dimension is very large dimension which contains 16000000 records.When I build cube ,this step( Build Dimension Dictionary ) would cost a very long time . could it be possible to build one time, so that other cube could use it directly rather than build after build. Another question is when the user dimension changes, each cube would be modified,it cost a lot,any suggestion ? -- Sent from: http://apache-kylin.74782.x6.nabble.com/
