[
https://issues.apache.org/jira/browse/HBASE-21007?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mike Drob reassigned HBASE-21007:
---------------------------------
Assignee: Bosko Devetak
Thanks for the patch, [~bdevetak]. I have added you to our "contributors"
group, so you should be able to self-assign issues in the future. I have
already assigned this one to you.
The change makes sense to me. Do you think it is possible to add a unit test to
verify this and prevent regressions in the future?
When submitting patches in the future, we encourage folks to use {{git
--format-patch}} so that they retain authorship information and we can properly
attribute credit. Alternatively, there is a helper script at
{{dev-support/submit-patch.py}}.
> Memory leak in HBase rest server
> --------------------------------
>
> Key: HBASE-21007
> URL: https://issues.apache.org/jira/browse/HBASE-21007
> Project: HBase
> Issue Type: Bug
> Components: REST
> Affects Versions: 1.4.0, 1.4.6
> Reporter: Bosko Devetak
> Assignee: Bosko Devetak
> Priority: Critical
> Attachments: HBASE-21007.001.patch
>
>
> When using the URIs like this:
>
> /sometable/*?limit=$limit&startrow=$startrow&endrow=$endrow
>
> where *$limit* is smaller than the range between *$startrow* and *$endrow*,
> the rest server will start leaking memory.
>
>
> The bug is in the *TableScanResource.java* class. Basically, the
> ResultScanner is not being closed in next() method when the limit has been
> reached.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
