As we were running 2 receivers and 2 senders (with receiver 1, sender 2 connected to Q1 and receiver 2, sender 1 connected to Q2), the receivers after some time (about 10 days of continuous running) stopped receiving messages (messages are there and could be seen using explorer) while both senders are able to send.
The receiver system is browsing the message first (doing some work) then receiving it.
The senders are still able to publish successfully. I enabled monitoring but received no monitor messages. We are the HA 7.3 router.
We are trying to re-produce the problem!
I know my statement is too general but any thoughts or pointers?
I don't think there was an out of memory exception. We are only printing the text of the exception, so I saw the timer exception but I didn't see any other exceptions meaning that there were no other exceptions reported.
SwiftMQ was idle for a couple of days. Is their a timer on the connection being idle?
There is a keepalive timer but that's only for the case when the phys. connection is down. In that case the connection is closed. So if you have a network problem, the connection is closed and you will not receive further messages. If you use a timer to send messages in a regular interval, for example, and the connection is closed, you will get an exception.
Anyway, these are all speculations. Log the whole stack trace and send it then we will know more. Since we don't have any issues nor open bugs, this is probably an application problem, e.g. use of sessions from multiple threads etc.