[ 
https://issues.apache.org/jira/browse/DRILL-3560?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14652831#comment-14652831
 ] 

Aman Sinha commented on DRILL-3560:
-----------------------------------

This works for me.. so I am not sure what was the scenario that you tried.  Let 
me describe the test that I ran: 
I have attached the table data that I used.  At the top level, it has 3 
directories:  1994, 1995, 1996.  Each of these has 4 subdirectories: 
Q1, Q2, Q3, Q4.   There is 1 data file within each of these subdirectories. 

Here's the EXPLAIN plan for the query:
{code}
0: jdbc:drill:zk=local> explain plan for select * from 
dfs.`data/multilevel/parquet` where dir0 = '1995';
+------+------+
| text | json |
+------+------+
| 00-00    Screen
00-01      Project(*=[$0])
00-02        Project(*=[$0])
00-03          Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath 
[path=file:/Users/asinha/data/multilevel/parquet/1995/Q1/orders_95_q1.parquet], 
ReadEntryWithPath 
[path=file:/Users/asinha/data/multilevel/parquet/1995/Q2/orders_95_q2.parquet], 
ReadEntryWithPath 
[path=file:/Users/asinha/data/multilevel/parquet/1995/Q3/orders_95_q3.parquet], 
ReadEntryWithPath 
[path=file:/Users/asinha/data/multilevel/parquet/1995/Q4/orders_95_q4.parquet]],
 selectionRoot=file:/Users/asinha/data/multilevel/parquet, numFiles=4, 
columns=[`*`]]])
{code}
Note that the Scan only shows 1995 directory, not the other years. 

Can you provide more details of your test case ?   Thanks. 

> Make partition pruning work for directory queries
> -------------------------------------------------
>
>                 Key: DRILL-3560
>                 URL: https://issues.apache.org/jira/browse/DRILL-3560
>             Project: Apache Drill
>          Issue Type: New Feature
>          Components: Query Planning & Optimization
>    Affects Versions: 1.1.0
>            Reporter: Stefán Baxter
>            Assignee: Aman Sinha
>
> Currently queries that include directory conditions are not optimized at all 
> and the directory expression (dir0 = 'something') is evaluated for every 
> record of every file for every directory.
> This could be optimized to fail directories and allow for the same kind of 
> partition pruning for directories as for other scenarios where data has been 
> partitioned.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to