[
https://issues.apache.org/jira/browse/ARTEMIS-4569?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17807989#comment-17807989
]
ASF subversion and git services commented on ARTEMIS-4569:
----------------------------------------------------------
Commit 2760a3e911f97ba940f378b4aacb40cd5ee30ebc in activemq-artemis's branch
refs/heads/main from Clebert Suconic
[ https://gitbox.apache.org/repos/asf?p=activemq-artemis.git;h=2760a3e911 ]
ARTEMIS-4569 Improving order in which the list is added versus cancel call
I just did some static analysis of this code, and I believe it would be better
to first add to the list
before setting the cancel task.
Reason for that is in case the Runnable is dequeued between the add in the
deQueue and setting the cancel task.
Possibility is remote but my OCD wouldn't let me ignore this small possibility.
> Blocked Producers Memory Leaks
> ------------------------------
>
> Key: ARTEMIS-4569
> URL: https://issues.apache.org/jira/browse/ARTEMIS-4569
> Project: ActiveMQ Artemis
> Issue Type: Improvement
> Affects Versions: 2.31.2
> Reporter: Clebert Suconic
> Assignee: Clebert Suconic
> Priority: Major
> Fix For: 2.32.0
>
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> Say you have paging configured as blocking. And you keep opening and closing
> connections that will be blocked while producing.
> These Runnables will accumulate until the number of messages goes back to
> negative again.
> Also these are holding Connections and other server objects.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
