activemq-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ARTEMIS-1495) Creating and destroying many queues could deadlock the broker
Date Wed, 08 Nov 2017 15:02:00 GMT

    [ https://issues.apache.org/jira/browse/ARTEMIS-1495?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16244100#comment-16244100
] 

ASF GitHub Bot commented on ARTEMIS-1495:
-----------------------------------------

Github user clebertsuconic commented on the issue:

    https://github.com/apache/activemq-artemis/pull/1650
  
    I have measured the lock version against the lock free and I don't see any difference.
    
    The code is supposedly single threaded most of the time.. for that reason using the JVM
lock didn't change much the perf.
    
    
    I will take the lock free version  you made though.. as it's extending the same logic
we have now.


> Creating and destroying many queues could deadlock the broker
> -------------------------------------------------------------
>
>                 Key: ARTEMIS-1495
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-1495
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>          Components: Broker
>            Reporter: Francesco Nigro
>         Attachments: output.log
>
>
> Running a JMS test with 100 Producers/100Consumers/100 queues sending each >=1 messages
could lead the broker to deadlock while waiting the paging cursor when the (core) clients
disconnect (eg AMQ222022: Timed out waiting for paging cursor to stop FutureLatch(latch=java.util.concurrent.CountDownLatch@415d8105[Count
= 1]) OrderedExecutor(tasks=[FutureLatch(latch=java.util.concurrent.CountDownLatch@415d8105[Count
= 1])]))



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message