[GitHub] [activemq-artemis] FrederikP commented on issue #2577: ARTEMIS-1604 Artemis deadlock using MQTT Protocol

[GitBox <gi...@apache.org>]

FrederikP commented on issue #2577: ARTEMIS-1604 Artemis deadlock using MQTT Protocol
URL: https://github.com/apache/activemq-artemis/pull/2577#issuecomment-472000777
 
 
   > just realized @FrederikP comments...
   > 
   > do you have an idea on what could be a test reproducing your behavior? Are you closing / opening consumers too fast or anything like that? it's Stomp, right?
   
   nope, no idea how to reproduce it. But @franz1981 already seems to be able to reproduce it locally using debug as he stated in his comment.
   
   I'm still going to summarize the usage of our artemis cluster because it might help: We took a three machine cluster online on friday and a couple of hours later one of the nodes ran into this dead lock. There are many mqtt and some stomp connections sending data and mostly stomp and some jms connections retrieving the data. The sending mqtt connections are under our control and should behave well. One of the stomp senders that is not under our control sends one item and then disconnects and opens a new connection for another one and so on (we are trying to get him to improve this). This actually might have triggered this dead lock, but I won't be able to verify.
   
   After the node crashed we had to roll back to the old activemq (non-artemis) for production. The issue didn't come up on our staging environment.
   
   If you have any other specific questions about our setup, let me know.
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services