You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@activemq.apache.org by "Robbie Gemmell (Jira)" <ji...@apache.org> on 2022/10/05 08:43:00 UTC

[jira] (ARTEMIS-3929) Changing OpenWireProtocolManager::clientIDs to a concurrent Hash Map

    [ https://issues.apache.org/jira/browse/ARTEMIS-3929 ]


    Robbie Gemmell deleted comment on ARTEMIS-3929:
    -----------------------------------------

was (Author: jira-bot):
Commit fc8a9b952c40ee190f7bfd9f15c0f4af12cb45c7 in activemq-artemis's branch refs/heads/2.21.0.jbossorg-x from Clebert Suconic
[ https://gitbox.apache.org/repos/asf?p=activemq-artemis.git;h=fc8a9b952c ]

ARTEMIS-3929 Improving OpenWire clientIDSet

I have seen this contention while I was testing ARTEMIS-3928

This does change any semantics and current tests should be enough to validate this change

(cherry picked from commit 568eb70fcda9737b38871a3bff1e111bad784091)

downstream: ENTMQBR-7051


> Changing OpenWireProtocolManager::clientIDs to a concurrent Hash Map
> --------------------------------------------------------------------
>
>                 Key: ARTEMIS-3929
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-3929
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>            Reporter: Clebert Suconic
>            Assignee: Clebert Suconic
>            Priority: Major
>             Fix For: 2.25.0
>
>
> as I was testing ARTEMIS-3928 I found this contention with the clientID hashMap, what was causing the entire broker to freeze when every client was failing.



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