[
https://issues.apache.org/jira/browse/AMQ-5684?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14384424#comment-14384424
]
Timothy Bish commented on AMQ-5684:
-----------------------------------
Chuck, I have pushed some fixes to ActiveMQ master which should show up in the
next SNAPSHOT build of 5.12 that attempt to address this. I've implemented
support for AMQP drain and in doing so found some places where we were not
handling incoming flow events correctly that could cause what you have been
seeing. I couldn't reproduce this using the AmqpNetLite tests after applying
all these fixes, can you validate this on your end. I am still seeing some odd
behaviour with the .NET tests when I use the 'jms' transformer but they seem
pretty happy when run against a broker that is using the 'raw' or 'native'
transformer, I'll try and track down what's going on there as well.
> AMQP: messages get stuck
> ------------------------
>
> Key: AMQ-5684
> URL: https://issues.apache.org/jira/browse/AMQ-5684
> Project: ActiveMQ
> Issue Type: Bug
> Components: AMQP
> Affects Versions: 5.12.0
> Environment: 5.12-SNAPSHOT 20150322 running on Fedora linux
> Reporter: Chuck Rolke
> Assignee: Timothy Bish
> Attachments: AMQ5684-reproducer.cs, second-receiver-hang-20150323.html
>
>
> While investigating AMQ-5646 the same test failed in a different way.
> It did not progress as far. The test:
> * Client opens a connection and a session.
> * Client opens a sender and sends 20 messages
> * Client opens a receiver with a credit of two
> The receiver receives one message and that's all. The expectation is to
> receive two messages.
> A trace to be attached.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
