[ https://issues.apache.org/activemq/browse/AMQ-736?page=comments#action_36231 ]
Danielius Jurna commented on AMQ-736: ------------------------------------- The problem is not in the broker, but in JournalMessageStore. Whet JoiurnalMessageStore is performing checkpoint, consumers doesn't receive messages during that checkpoint. The bug is that during checkpoint, messages are removed from cpAddedMessageIds list, and as I understand, this is temporal message store used during checkpoint. And during checkpoint messages shouldn't be removed from there, because consumer is using this member to get messages, which are already removed from short term storage, but not yet added to long term storage. So fix is preatty simple. I'm adding the patch against the main trunk of ActiveMQ. Hope some ActiveMQ developers will take a look. > Broker is not delivering all messages to slow consumer > ------------------------------------------------------ > > Key: AMQ-736 > URL: https://issues.apache.org/activemq/browse/AMQ-736 > Project: ActiveMQ > Type: Bug > Components: Broker > Versions: 4.0 > Environment: Linux/Windows Java 1.5.0_06 > Reporter: Danielius Jurna > Priority: Critical > Attachments: SlowConsumerTest.java, slowConsumer.patch > > > When producer produces more messages than consumer can handle, after some > time some messages are not delivered to the consumer (they are just skipped). > To get those messages from the broker, you need to reconnect the consumer. It > seems that this bug only occures if message persistance is turned on in the > broker. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: https://issues.apache.org/activemq/secure/Administrators.jspa - For more information on JIRA, see: http://www.atlassian.com/software/jira
