[
https://issues.apache.org/jira/browse/CURATOR-462?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16432573#comment-16432573
]
ASF GitHub Bot commented on CURATOR-462:
----------------------------------------
GitHub user krajcsovszkig-ms opened a pull request:
https://github.com/apache/curator/pull/263
[CURATOR-462] InterProcessSemaphoreV2 leaves orphaned lease node if
acquiring the semaphore is interrupted at the wrong time
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/krajcsovszkig-ms/curator CURATOR-462
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/curator/pull/263.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #263
----
commit f9128a87c837fd4a687c3d54c73ea779e46085fc
Author: krajcsovszkig-ms <krajcsovszkig-ms@...>
Date: 2018-04-10T12:08:45Z
Fix CURATOR-462 -- return lease created in
org.apache.curator.framework.recipes.locks.InterProcessSemaphoreV2.internalAcquire1Lease(Builder<Lease>,
long, boolean, long) when the wait for it to become active is interrupted;
test for the fix
----
> InterProcessSemaphoreV2 leaves orphaned lease node if acquiring the semaphore
> is interrupted at the wrong time
> --------------------------------------------------------------------------------------------------------------
>
> Key: CURATOR-462
> URL: https://issues.apache.org/jira/browse/CURATOR-462
> Project: Apache Curator
> Issue Type: Bug
> Components: Recipes
> Affects Versions: 2.11.0, 3.2.0, 3.2.1, 2.11.1, 3.3.0, 2.12.0, 4.0.0, 4.0.1
> Reporter: Gergely Krajcsovszki
> Priority: Major
>
> Since CURATOR-315, lease nodes created in
> org.apache.curator.framework.recipes.locks.InterProcessSemaphoreV2.internalAcquire1Lease(Builder<Lease>,
> long, boolean, long) are not returned if the wait for them to become active
> is interrupted. This is because the lease node is now only added to the
> builder at the end of the method, so the caller doesn't know about this node
> and cannot close it when catching an exception.
> We have already fixed this locally in our own dependent library so I'll make
> a PR with our fix and a test soon. We are still on the 2.x branch so we'd
> like if this fix could be in 2.13.0.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
