[jira] [Closed] (ARTEMIS-4432) openwire connection failure handling is bypassing the actor and ignoring the operation context leading to contention in error
[ https://issues.apache.org/jira/browse/ARTEMIS-4432?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Clebert Suconic closed ARTEMIS-4432. Resolution: Fixed > openwire connection failure handling is bypassing the actor and ignoring the > operation context leading to contention in error > - > > Key: ARTEMIS-4432 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4432 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: OpenWire >Affects Versions: 2.30.0 >Reporter: Gary Tully >Assignee: Gary Tully >Priority: Major > Fix For: 2.32.0 > > > On a kill -9 test, for a transacted batch openwire consumer, it is possibly > for a batch to get stuck in delivery due to contention on the current > transaction. A restart is needed to have redeliver happen in this case. > The contention can be avoided by respecting the ordered actor and operation > context in the fail handling that originates from the netty handler when a > connection dies due to socket error. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Closed] (ARTEMIS-4432) openwire connection failure handling is bypassing the actor and ignoring the operation context leading to contention in error
[ https://issues.apache.org/jira/browse/ARTEMIS-4432?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Clebert Suconic closed ARTEMIS-4432. > openwire connection failure handling is bypassing the actor and ignoring the > operation context leading to contention in error > - > > Key: ARTEMIS-4432 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4432 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: OpenWire >Affects Versions: 2.30.0 >Reporter: Gary Tully >Assignee: Gary Tully >Priority: Major > Fix For: 2.32.0 > > > On a kill -9 test, for a transacted batch openwire consumer, it is possibly > for a batch to get stuck in delivery due to contention on the current > transaction. A restart is needed to have redeliver happen in this case. > The contention can be avoided by respecting the ordered actor and operation > context in the fail handling that originates from the netty handler when a > connection dies due to socket error. -- This message was sent by Atlassian Jira (v8.20.10#820010)
