carbondata-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "QiangCai (JIRA)" <>
Subject [jira] [Created] (CARBONDATA-782) Support SORT_COLUMNS
Date Thu, 16 Mar 2017 03:34:42 GMT
QiangCai created CARBONDATA-782:

             Summary: Support SORT_COLUMNS
                 Key: CARBONDATA-782
             Project: CarbonData
          Issue Type: New Feature
            Reporter: QiangCai
            Assignee: QiangCai

The tasks of SORT_COLUMNS:

1.Support create table with sort_columns property.
e.g. tblproperties('sort_columns' = 'col7,col3')
The table with SORT_COLUMNS property will be sorted by SORT_COLUMNS. The order of columns
is decided by SORT_COLUMNS.

2.Change the encoding rule of SORT_COLUMNS
Firstly, the rule of column encoding will keep consistent with previous.
Secondly, if a column of SORT_COLUMNS is a measure before, now this column will be created
as a dimension. And this dimension is a no-dicitonary column(Better to use other direct-dictionary).
Thirdly, the dimension of SORT_COLUMNS have RLE and ROWID page, other dimension have only
RLE(not sorted).

3.The start/end key should be composed of SORT_COLUMNS.
Using SORT_COLUMNS to build start/end key during data loading and select query.

This message was sent by Atlassian JIRA

View raw message