You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Robert Hou (JIRA)" <ji...@apache.org> on 2016/12/01 23:58:58 UTC
[jira] [Commented] (DRILL-5093) Explain plan shows all partitions
when query scans all partitions, and filter pushdown is used with metadata
caching.
[ https://issues.apache.org/jira/browse/DRILL-5093?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15713469#comment-15713469 ]
Robert Hou commented on DRILL-5093:
-----------------------------------
The two tables have the same data.
> Explain plan shows all partitions when query scans all partitions, and filter pushdown is used with metadata caching.
> ---------------------------------------------------------------------------------------------------------------------
>
> Key: DRILL-5093
> URL: https://issues.apache.org/jira/browse/DRILL-5093
> Project: Apache Drill
> Issue Type: Bug
> Components: Query Planning & Optimization
> Affects Versions: 1.9.0
> Reporter: Robert Hou
> Assignee: Jinfeng Ni
> Attachments: 0_0_1.parquet, 0_0_2.parquet, 0_0_3.parquet, 0_0_4.parquet, 0_0_5.parquet, drill.parquet_metadata
>
>
> This query scans all the partitions because the partitions cannot be pruned. When metadata caching is used, the explain plan shows all the partitions, when it should only show the parent.
> 0: jdbc:drill:zk=10.10.100.186:5181/drill/rho> explain plan for select \* from orders_parts_metadata;
> +------+------+
> | text | json |
> +------+------+
> | 00-00 Screen
> 00-01 Project(*=[$0])
> 00-02 Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath [path=/drill/testdata/filter/orders_parts_metadata/0_0_1.parquet], ReadEntryWithPath [path=/drill/testdata/filter/orders_parts_metadata/0_0_3.parquet], ReadEntryWithPath [path=/drill/testdata/filter/orders_parts_metadata/0_0_4.parquet], ReadEntryWithPath [path=/drill/testdata/filter/orders_parts_metadata/0_0_5.parquet], ReadEntryWithPath [path=/drill/testdata/filter/orders_parts_metadata/0_0_2.parquet]], selectionRoot=/drill/testdata/filter/orders_parts_metadata, numFiles=5, usedMetadataFile=true, cacheFileRoot=/drill/testdata/filter/orders_parts_metadata, columns=[`*`]]])
> To reproduce the problem, put the attached files into a directory. Then create the metadata:
> refresh table metadata dfs.`path_to_directory`;
> For example, if you put the files in /drill/testdata/filter/orders_parts_metadata, then run this sql command
> refresh table metadata dfs.`/drill/testdata/filter/orders_parts_metadata`;
> Here is the same query with a table that does not have metadata caching.
> 0: jdbc:drill:zk=10.10.100.186:5181/drill/rho> explain plan for select \* from orders_parts;
> +------+------+
> | text | json |
> +------+------+
> | 00-00 Screen
> 00-01 Project(*=[$0])
> 00-02 Scan(groupscan=[ParquetGroupScan [entries=[ReadEntryWithPath [path=maprfs:///drill/testdata/filter/orders_parts]], selectionRoot=maprfs:/drill/testdata/filter/orders_parts, numFiles=1, usedMetadataFile=false, columns=[`*`]]])
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)