Re: kylin cube dimension 改动问题

2017-05-19 Thread Billy Liu
I may made some mistakes in my above reply. I don't remember every detail
very clearly. If the cube has existing segments, it's not allowed to edit.
You have to clone the cube and edit it.

2017-05-19 16:16 GMT+08:00 elicer :

> For answer2:If I modify the cube and continue the cube build only the
> incremental data will be affected right. If I want the old data to reflect
> this field(This is a existing field, but previous not selected as one of
> the
> dimension), I need to rebuild the while cute right.
>
> --
> View this message in context: http://apache-kylin.74782.x6.
> nabble.com/kylin-cube-dimension-tp8033p8040.html
> Sent from the Apache Kylin mailing list archive at Nabble.com.
>


Re: kylin cube dimension 改动问题

2017-05-19 Thread elicer
For answer2:If I modify the cube and continue the cube build only the
incremental data will be affected right. If I want the old data to reflect
this field(This is a existing field, but previous not selected as one of the
dimension), I need to rebuild the while cute right.

--
View this message in context: 
http://apache-kylin.74782.x6.nabble.com/kylin-cube-dimension-tp8033p8040.html
Sent from the Apache Kylin mailing list archive at Nabble.com.


Re: Re: kylin cube dimension 改动问题

2017-05-19 Thread gaolv123...@163.com
同问



gaolv123...@163.com
 
发件人: Billy Liu
发送时间: 2017-05-19 15:05
收件人: dev
主题: Re: kylin cube dimension 改动问题
A few suggestions:
1. Use View instead Table, View could help you isolate the underlying table
changes and keep the data model stable.
2. If column added only, you could sync the table, edit the cube and
continue build the new segment.
 
 
在 2017年5月19日 下午2:42,elicer 写道:
 
> Hi All:
> 如果大家的报表字段改动比较平凡的话,有没什么好的方法避免重新去create cube and  rebuid cube,数据量大,开小太大啊。
>
> --
> View this message in context: http://apache-kylin.74782.x6.
> nabble.com/kylin-cube-dimension-tp8033.html
> Sent from the Apache Kylin mailing list archive at Nabble.com.
>


Re: kylin cube dimension 改动问题

2017-05-19 Thread Billy Liu
A few suggestions:
1. Use View instead Table, View could help you isolate the underlying table
changes and keep the data model stable.
2. If column added only, you could sync the table, edit the cube and
continue build the new segment.


在 2017年5月19日 下午2:42,elicer 写道:

> Hi All:
> 如果大家的报表字段改动比较平凡的话,有没什么好的方法避免重新去create cube and  rebuid cube,数据量大,开小太大啊。
>
> --
> View this message in context: http://apache-kylin.74782.x6.
> nabble.com/kylin-cube-dimension-tp8033.html
> Sent from the Apache Kylin mailing list archive at Nabble.com.
>