Steven Youtsey created NIFI-6391:
------------------------------------
Summary: ConsumeJMS/PublishJMS - Excessive connections created
when get/publish operations continually fail
Key: NIFI-6391
URL: https://issues.apache.org/jira/browse/NIFI-6391
Project: Apache NiFi
Issue Type: Bug
Components: Extensions
Affects Versions: 1.8.0
Reporter: Steven Youtsey
PublishJMS repeatedly threw "ResourceAllocationException: too many open
connections" after successive publish failures due to issues on the remote JMS
Broker. The cause of this is the specification and implementation of the
JMSConnectionFactoryProviderDefinition.resetConnectionFactory(ConnectionFactory
cf) method. Setting the ConnectionFactory to null may well indeed close
connections upon 'destruction', but it may take awhile for the GC to run; in
the meantime, more connections are opened. The connections need to be manually
closed rather than waiting for a GC.
>From AbstractJMSProcessor.onTrigger(), need to call worker.shutdown() prior to
>resetConnectionFactory().
Also, noticed some problems iwth the ConnectionFactoryProviderDefinition
implementatoins wrt the resetConnectionFactory methods. The factory is nulled
but never re-initialized but for onEnabled(); which will lead to a NPE at some
point.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
