[
https://issues.apache.org/jira/browse/DRILL-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15746676#comment-15746676
]
ASF GitHub Bot commented on DRILL-5121:
---------------------------------------
Github user jinfengni commented on the issue:
https://github.com/apache/drill/pull/690
The proposed change looks fine for case insensitive store. On the other
hand, have you given some thoughts of @Ben-Zvi 's comment about case sensitive
store? Will it work for case sensitive store?
You may add one simple testcase to verify that the fix works as expected.
> A memory leak is observed when exact case is not specified for a column in a
> filter condition
> ---------------------------------------------------------------------------------------------
>
> Key: DRILL-5121
> URL: https://issues.apache.org/jira/browse/DRILL-5121
> Project: Apache Drill
> Issue Type: Bug
> Components: Execution - Relational Operators
> Affects Versions: 1.6.0, 1.8.0
> Reporter: Karthikeyan Manivannan
> Assignee: Karthikeyan Manivannan
> Fix For: Future
>
> Original Estimate: 2h
> Remaining Estimate: 2h
>
> When the query SELECT XYZ from dfs.`/tmp/foo' where xYZ like "abc", is
> executed on a setup where /tmp/foo has 2 Parquet files, 1.parquet and
> 2.parquet, where 1.parquet has the column XYZ but 2.parquet does not, then
> there is a memory leak.
> This seems to happen because xYZ seem to be treated as a new column.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
