[
https://issues.apache.org/jira/browse/HIVE-14181?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16095376#comment-16095376
]
Aihua Xu commented on HIVE-14181:
---------------------------------
[~szita] That is a challenge to keep HDFS async with metadata during insertion
and deletion. Agree that we can print some kind of warning to the user at least
and let the user clean up the data manually.
Another thought: when HDFS trash is turned off, seems throwing the warning is
what we can do; when HDFS trash is turned on, since we can recover HDFS files,
then we can keep HDFS and metadata async by recovering HDFS file and reverting
metadata change if anything fails.
How do you think?
> DROP TABLE in hive doesn't Throw Error
> --------------------------------------
>
> Key: HIVE-14181
> URL: https://issues.apache.org/jira/browse/HIVE-14181
> Project: Hive
> Issue Type: Bug
> Environment: Hive 1.1.0
> CDH 5.5.1-1
> Reporter: Pranjal Singh
> Assignee: Adam Szita
> Labels: easyfix
> Attachments: HIVE-14181.1.patch, HIVE-14181.2.patch
>
>
> drop table table_name doen't throw an error if drop table fails.
> I was dropping a table and my trash didn't have enough space to hold the
> table but the drop table command showed success and the table wasn't deleted.
> But the hadoop fs -rm -r /hive/xyz.db/table_name/ gave an error "Failed to
> move to trash" because I didnot have enough space quota in my trash.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
