You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@kylin.apache.org by 高扬02 <ga...@bianfeng.com> on 2019/07/15 07:06:36 UTC

答复: Where the high cardinality dimension should be placed in the rowkey

Hi Temple Zhou,
Thank you for your reply.
Yes, the previous dimensions have filtering requirements like “where xx in (xx,xx)”.
Thanks very much.

发件人: Temple Zhou [mailto:dbaztp@gmail.com]
发送时间: 2019年7月15日 14:59
收件人: user <us...@kylin.apache.org>
主题: Re: Where the high cardinality dimension should be placed in the rowkey

If those dimensions before the two with high cardinality will be used for filtering, maybe, you are right.
[image.png]

On Mon, Jul 15, 2019 at 2:49 PM 高扬02 <ga...@bianfeng.com>> wrote:
Hello,
I have two high cardinality dimensions , which only “group by bmid_user”, but no “where bmid_user in (xx,xx)”, so I put them in the end of rowkeys. Is this right?
[cid:image003.png@01D53B1E.C83D83F0]

Thanks very much.