]
Stefano Maestri moved JBEAP-14759 to WFLY-10414:
------------------------------------------------
Project: WildFly (was: JBoss Enterprise Application Platform)
Key: WFLY-10414 (was: JBEAP-14759)
Workflow: GIT Pull Request workflow (was: CDW with loose statuses v1)
Component/s: JCA
(was: JCA)
Affects Version/s: 13.0.0.Beta1
12.0.0.Final
(was: 7.1.2.GA)
(was: 7.2.0.CD13.GA)
Cached-connection-manager DEBUG connection closure could be too eager
with enlistment-trace==false
--------------------------------------------------------------------------------------------------
Key: WFLY-10414
URL:
https://issues.jboss.org/browse/WFLY-10414
Project: WildFly
Issue Type: Bug
Components: JCA
Affects Versions: 13.0.0.Beta1, 12.0.0.Final
Reporter: Stefano Maestri
Assignee: Stefano Maestri
Labels: gss
There is some corner cases in which IJ isn't able to avoid IJ000100 (closing
connection for you) on connection already marked as invalid and firing IJ000100 closing
it, even if the connection would be closed as invalid immediately after. It doesn't
create problems to the connection pool, but the redundant WARN isn't useful and could
be read as connection leak.
It happens only with enlistment-trace=false (which is default since 7.2) set for the
involved datasource