[
https://issues.apache.org/jira/browse/NIFI-16287?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Joe Witt updated NIFI-16287:
----------------------------
Resolution: Fixed
Status: Resolved (was: Patch Available)
> Ensure orderly shutdown of active and idle JMS workers
> ------------------------------------------------------
>
> Key: NIFI-16287
> URL: https://issues.apache.org/jira/browse/NIFI-16287
> Project: Apache NiFi
> Issue Type: Improvement
> Reporter: Pierre Villard
> Assignee: Pierre Villard
> Priority: Major
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> AbstractJMSProcessor currently tracks only idle workers in its pool, so
> workers executing ConsumeJMS or PublishJMS may retain JMS connections when a
> processor is unscheduled or loses primary-node ownership. Introduce
> scheduling-cycle ownership for all JMS workers, close active and idle workers
> during early lifecycle shutdown, prevent retired workers from being rebuilt
> or returned to the pool, and make worker shutdown idempotent and
> exception-tolerant. Add deterministic tests covering concurrent construction
> and shutdown, blocked receives, delayed acknowledgement and redelivery,
> primary-node transitions, and interrupted publishes.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
