[jira] [Commented] (DRILL-1115) parquet writer never finishes when the source contains huge number of small files

["Rahul Challapalli (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DRILL-1115?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14056956#comment-14056956 ] 

Rahul Challapalli commented on DRILL-1115:
------------------------------------------

As jacques pointed out drill seems to be consuming all the memory and once it reaches this stage it could be going into an infinite loop or something of that sort

> parquet writer never finishes when the source contains huge number of small files
> ---------------------------------------------------------------------------------
>
>                 Key: DRILL-1115
>                 URL: https://issues.apache.org/jira/browse/DRILL-1115
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Storage - Writer
>            Reporter: Rahul Challapalli
>
> git.commit.id.abbrev=790a2ad
> Build # 26246
> When we try to use 'create table as.....' and the source folder contains around 5000 text files, drill never completes. I left the query to tun overnight, but it still didn't complete. However I see nonstop activity in the log files which suggests drill is actually doing something.
> Cluster Size : 2
> Each file contains only a single number.



--
This message was sent by Atlassian JIRA
(v6.2#6252)