You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Hanifi Gunes (JIRA)" <ji...@apache.org> on 2014/10/21 01:45:33 UTC

[jira] [Commented] (DRILL-1547) Selecting records with multiple null fields fails with AssertionError

    [ https://issues.apache.org/jira/browse/DRILL-1547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14177664#comment-14177664 ] 

Hanifi Gunes commented on DRILL-1547:
-------------------------------------

The patch enforces writers to explicitly check for buffer bounds to avoid IndexOutOfBounds errors & makes writer hierarchy to stop immediately in case of a write error.

> Selecting records with multiple null fields fails with AssertionError
> ---------------------------------------------------------------------
>
>                 Key: DRILL-1547
>                 URL: https://issues.apache.org/jira/browse/DRILL-1547
>             Project: Apache Drill
>          Issue Type: Bug
>            Reporter: Hanifi Gunes
>            Assignee: Hanifi Gunes
>         Attachments: DRILL-1547.1.patch.txt
>
>
> If reader hits a records with multiple `null` valued fields whose vectors are not yet allocated, running capacity check twice throws AssertionError. 
> Take
> {panel:title=a.json}
> \{"some":null, "field": null\}
> {panel}
> Running
> {code:sql}
> select * from dfs.`a.json`
> {code}
> fails with AssertionError.
> The work around is to check value capacity only if vector is allocated. This is an issue that affects sparse JSON files.



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