[
https://issues.apache.org/jira/browse/DRILL-6863?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16694543#comment-16694543
]
Denys Ordynskiy commented on DRILL-6863:
----------------------------------------
[~ben-zvi] this error in the log file is expected since I haven't permission to
rename folders on S3. When Drill is trying to drop table, one of the steps is
rename table. The main issue here is that drill returns successfully response
on drop table query, but it should returns an access error like in logs.
I updated expected result in jira description.
> Drop table is not working on amazon S3
> --------------------------------------
>
> Key: DRILL-6863
> URL: https://issues.apache.org/jira/browse/DRILL-6863
> Project: Apache Drill
> Issue Type: Bug
> Affects Versions: 1.15.0
> Reporter: Denys Ordynskiy
> Assignee: Bohdan Kazydub
> Priority: Major
> Fix For: 1.15.0
>
>
> Parquet table was created using CTAS on S3.
> Request "drop table
> s3.tmp.`/drill/transitive_closure/DRILL_6173_filterPushdown/tab1`"
> returns successfully response:
> "Table [/drill/transitive_closure/DRILL_6173_filterPushdown/tab1] dropped"
>
> *Actual result:*
> Drill did not delete the table `tab1`, files still present on S3 storage.
> In "drillbit.out":
> {code:java}
> 23:54:49.661 [2416095b-6544-fc80-0dfa-2fc19c4dee0e:foreman] ERROR
> o.apache.hadoop.fs.s3a.S3AFileSystem - rename: src not found
> /drill/transitive_closure/DRILL_6173_filterPushdown/tab1
> {code}
>
> *Expected result:*
> Drill should delete table from S3 storage.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
