You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Victoria Markman (JIRA)" <ji...@apache.org> on 2015/01/21 20:39:36 UTC

[jira] [Commented] (DRILL-2045) NPE when querying INFORMATION_SCHEMA table

    [ https://issues.apache.org/jira/browse/DRILL-2045?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14286140#comment-14286140 ] 

Victoria Markman commented on DRILL-2045:
-----------------------------------------

It does not look like an upgrade issue (even though I did not follow suggested upgrade path in our documentation)
I examined zookeeper's data related to drill's storage plugins and I found out that metadata related to views was wiped out, when view files where still on disk.
No sure, how this happened yet. When I deleted view file on disk, select from INFORMATION_SCHEMA.`TABLES` worked.

> NPE when querying INFORMATION_SCHEMA table
> ------------------------------------------
>
>                 Key: DRILL-2045
>                 URL: https://issues.apache.org/jira/browse/DRILL-2045
>             Project: Apache Drill
>          Issue Type: Bug
>    Affects Versions: 0.8.0
>            Reporter: Victoria Markman
>         Attachments: drill-2045.log
>
>
> I'm filing this bug as a placeholder to investigate the reason of NPE on
> "select * from INFORMATION_SCHEMA.`TABLES`" on my cluster.
> My strong suspicion is that it is related to the way I upgrade drill on a cluster and this is somehow scrambles zookeeper's state. 
> My steps are:
>     * stop drillbits
>     * uninstall drill rpm: "rpm -e <>"
>     * install a new rpm: "rpm -i <>"
>     * service mapr-warden restart (I don't stop warden before an upgrade)



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