[
https://issues.apache.org/jira/browse/ARTEMIS-1271?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16100300#comment-16100300
]
ASF subversion and git services commented on ARTEMIS-1271:
----------------------------------------------------------
Commit c3eeea491053eff396ed4d5b55b1fdb3c8568639 in activemq-artemis's branch
refs/heads/1.x from [~martyntaylor]
[ https://git-wip-us.apache.org/repos/asf?p=activemq-artemis.git;h=c3eeea4 ]
ARTEMIS-1271 FileSystemMonitor locking improvement
(cherry picked from commit ce756d46f9c90e4d7d0eb0e0090384d6d646af3e)
> Deadlock on shutdown may occur if using shared store and network cable is
> pulled
> --------------------------------------------------------------------------------
>
> Key: ARTEMIS-1271
> URL: https://issues.apache.org/jira/browse/ARTEMIS-1271
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Affects Versions: 2.1.0
> Reporter: Martyn Taylor
> Assignee: Martyn Taylor
> Fix For: 1.5.6, 2.2.0
>
>
> The FileStoreMonitor locks it's monitor lock when scanning the file system
> for usage. If using a shared store and the network cable is pulled. The
> call to get file system usage may hang. This prevents the server from
> shutting down. There is a shutdownNow call on the relevant thread pool which
> will interupt the thread, but the stop() call is blocked on the
> FileStoreMonitor monitor lock.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
