Dequeue stops to function properly

[Qian Su <qs...@linkedin.com>]

Hi,

We've been using Activemq version 4.1.1 for almost a year, over TCP
transport, with journal + mysql DB for persistence. 

Recently, we observe in STG env. that some messages fail to be dequeued
properly from some queues, though consumers appear to function normally.
Some of these queues have existed for a long time and have been behaving
normal both in STG and prod. environments.  

The only recent changes we are aware of include the addition of 2 queues and
10 corresponding consumers, which bring us to a total of 10 queues (on one
broker machine) and ~40 consumers.

The volume of traffic is very low (<100 messages when the problem is
observed) in the staging environment. Through the jmx console interface, we
can see the delta between enqueue count and dequeue count and from whatever
reason the dequeue count ceases to increment.  Furthermore, when we try to
purge messages in the queue, they don't seem to be purged and instead, the
consumers are awakened, but still no messages processed.  

The only way we are aware of to bring the queues back to normal is to bring
down the broker. reinit mysql DB, and bring back the broker.

Does anybody have any directions as to what might have triggered the
unstable behavior of the queues? Can it be that we are reaching the limit of
number of queues per server? Or maybe some synchronization issue with the
journal?

Thanks,
Qian 
-- 
View this message in context: http://www.nabble.com/Dequeue-stops-to-function-properly-tp16872420s2354p16872420.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.