[
https://issues.apache.org/jira/browse/HIVE-19201?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16640386#comment-16640386
]
Nishant Bangarwa commented on HIVE-19201:
-----------------------------------------
[~Korge] Make sure the column names in druid are all in lowercase. If you have
a column in druid which has upper case characters and hive sends the query with
lowercase column names, it can lead to getting null in results.
> Hive doesn't read Druid data correctly
> --------------------------------------
>
> Key: HIVE-19201
> URL: https://issues.apache.org/jira/browse/HIVE-19201
> Project: Hive
> Issue Type: Bug
> Components: Druid integration, Hive
> Affects Versions: 2.3.0
> Environment: Ubuntu 16.4 TLS Desktop
> Druid 0.12.0 (StandAlone - Quickstart conf)
> Hive 2.3.0 (StandAlone - Quickstart conf)
> Also have Hadoop and Zookeeper running
> Reporter: Tournadre
> Priority: Blocker
> Attachments: DruidIntegration.tar
>
>
> I created an external table on hive pointing at my datasource : wikiticker on
> druid (I already defined the broker adress). However few colums appear as
> NULL without knowing why.
> I have also many time *.lck file in the metadatadb on Hive and Druid to
> delete (otherwise, queries don't work)
> So the describe statement lists the column and their type correctly, however
> I only get "NULL" when I query some column (string type). I checked in my
> Druid, the data are still here.
>
> Help ! :(
> PS : Sorry first issue and JIRA, didn't see my issue resolved yet.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
