You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Bohdan Kazydub (Jira)" <ji...@apache.org> on 2020/06/30 16:37:00 UTC
[jira] [Commented] (DRILL-7760) REFRESH TABLE METADATA increases
the planning time while querying over parquet files.
[ https://issues.apache.org/jira/browse/DRILL-7760?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17148815#comment-17148815 ]
Bohdan Kazydub commented on DRILL-7760:
---------------------------------------
[~mtewathia99] it's hard to tell anything without a query and table schema. Can you provide reproducible example with steps to reproduce, please?
> REFRESH TABLE METADATA increases the planning time while querying over parquet files.
> -------------------------------------------------------------------------------------
>
> Key: DRILL-7760
> URL: https://issues.apache.org/jira/browse/DRILL-7760
> Project: Apache Drill
> Issue Type: Bug
> Components: Metadata
> Affects Versions: 1.17.0
> Environment: Apache Drill Web UI ( 1.170.0 )
> Reporter: Mukul Tewathia
> Priority: Major
> Fix For: Future
>
>
> After running
> REFRESH TABLE METADATA <table>
> The planning time increases for a particular query increases from 35 secs to more than 3 minutes.
> Parquet file consist of 10M rows and 7 columns and is partitioned on 2 columns ( both the column contains 8 unique values ).
> What could be the possible explanation for this behaviour ?
--
This message was sent by Atlassian Jira
(v8.3.4#803005)