[
https://issues.apache.org/jira/browse/LENS-714?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14679637#comment-14679637
]
Hudson commented on LENS-714:
-----------------------------
UNSTABLE: Integrated in Lens-Commit #868 (See
[https://builds.apache.org/job/Lens-Commit/868/])
LENS-714 : Replace dimension tables with subquery in Columnar Rewriter
(amareshwari: rev 7610abaf951ca2c61e51b8941643c57f4bdc2f70)
*
lens-driver-jdbc/src/main/java/org/apache/lens/driver/jdbc/ColumnarSQLRewriter.java
* lens-driver-jdbc/src/test/java/org/apache/lens/driver/jdbc/TestJDBCFinal.java
*
lens-driver-jdbc/src/test/java/org/apache/lens/driver/jdbc/TestColumnarSQLRewriter.java
> Replace dimension tables with subquery in Columnar Rewriter
> -----------------------------------------------------------
>
> Key: LENS-714
> URL: https://issues.apache.org/jira/browse/LENS-714
> Project: Apache Lens
> Issue Type: Improvement
> Components: driver-jdbc
> Reporter: Sushil Mohanty
> Assignee: Sushil Mohanty
> Fix For: 2.3
>
> Attachments: LENS-714-2.patch
>
>
> In order to optimise the JDBC sql queries, we can skip directly joining with
> the dimension tables. Instead we can create a subquery with only required
> column for a table used in the query. For eg. If column id and name are
> queried from table dim1 the query can be rewritten as
> ... inner join (select id,name from dim1) as d1 instead of ...inner join dim1
> as d1.
> For dimension tables with descent size this helps in improving performance to
> a great extent. Planning to add this in ColumnarRewriter so that columnar DBs
> can leverage this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
