[ 
https://issues.apache.org/jira/browse/AMQ-9823?focusedWorklogId=997110&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-997110
 ]

ASF GitHub Bot logged work on AMQ-9823:
---------------------------------------

                Author: ASF GitHub Bot
            Created on: 19/Dec/25 13:21
            Start Date: 19/Dec/25 13:21
    Worklog Time Spent: 10m 
      Work Description: cshannon commented on PR #1566:
URL: https://github.com/apache/activemq/pull/1566#issuecomment-3675068855

   The failure was caused by a port issue which happens. It looks like we are 
no longer using the re-run failed tests flag which I feel like is kind of 
required for projects like this just because it's hard to avoid the following 
types of issues (ie ports in use) 
   
   ```
   [ERROR]   
JMSConsumerTest>CombinationTestSupport.runBare:107->testPrefetch1MessageNotDispatched:803
   [ERROR] Errors: 
   [ERROR]   
ActiveMQXASslConnectionFactoryTest>CombinationTestSupport.runBare:113->CombinationTestSupport.runBare:107->ActiveMQSslConnectionFactoryTest.testCreateFailoverTcpConnectionUsingKnownPort:106->ActiveMQSslConnectionFactoryTest.createBroker:339
 ยป IO Failed to bind to server socket: 
tcp://localhost:61610?wireFormat.tcpNoDelayEnabled=true due to: 
java.net.BindException: Address already in use
   ```
   
   @mattrpav - The other interesting thing is both my recent PRs jenkins is not 
finding the test output so it's not showing the failed tests. I have to 
manually dig through the output to find the failed test. I'm wondering if the 
parallel test stuff is causing issues with jenkins finding the file.
   
   The output shows: `No test report files were found. Configuration error?`




Issue Time Tracking
-------------------

    Worklog Id:     (was: 997110)
    Time Spent: 0.5h  (was: 20m)

> KahaDBStore ackAndPreparedMap is not properly cleared when recovered acks are 
> removed 
> --------------------------------------------------------------------------------------
>
>                 Key: AMQ-9823
>                 URL: https://issues.apache.org/jira/browse/AMQ-9823
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker, KahaDB
>    Affects Versions: 6.2.0, 5.19.1, 6.1.8
>            Reporter: Christopher L. Shannon
>            Assignee: Christopher L. Shannon
>            Priority: Major
>             Fix For: 6.3.0, 5.19.2, 6.2.1
>
>          Time Spent: 0.5h
>  Remaining Estimate: 0h
>
> Kahadb tracks prepared acks for an XA transaction to be recovered inside of 
> the 
> [ackedAndPreparedMap|https://github.com/apache/activemq/blob/7aa5bad9e3055f65b15f047ceb9f4f24327f5309/activemq-kahadb-store/src/main/java/org/apache/activemq/store/kahadb/KahaDBStore.java#L416].
>   The map contains the set of prepared acks that were recovered. After 
> commit/rollback the acks are supposed to be 
> "[forgotten|https://github.com/apache/activemq/blob/7aa5bad9e3055f65b15f047ceb9f4f24327f5309/activemq-kahadb-store/src/main/java/org/apache/activemq/store/kahadb/KahaDBStore.java#L458]";
>  and removed from the map. 
> They method correctly clears from the set but there is a bug where an empty 
> set is left in the map because the map is 
> [checked|https://github.com/apache/activemq/blob/7aa5bad9e3055f65b15f047ceb9f4f24327f5309/activemq-kahadb-store/src/main/java/org/apache/activemq/store/kahadb/KahaDBStore.java#L468]
>  to see if it is empty vs checking the set so the set never gets removed so 
> this could in theory have a small memory leak. In practice it's probably not 
> much of an issue but worth fixing as it's one line.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
For further information, visit: https://activemq.apache.org/contact


Reply via email to