Nominal wrote:
No problem, thanks for the reply. We went ahead and implemented a
Thrift based daemon to maintain a connection to the qpidd daemon.
You now use a single connection for all messages sent? That takes cares
of the large number of failover exchange subscription queues, right?
However we are still seeing the same memory consumption issue. As I
have mentioned, the issue only arises when the qpid daemon is
operating in cluster mode. Is there any place I can look for
instruction on creating tests to assert what I am observing for the
purpose of providing feedback? It seems that even with the
less-than-ideal use pattern we have implemented, qpid should still
not exhaust all the memory on a system.
You are observing the memory usage continually going up until all memory
is exhausted? Are messages being dequeued from the queues? What does
'qpid-stat -q' show? (Thats a tool available with the python client,
also in svn under qpid/python/commands).
Also if you can point me in
the right direction for the cluster code that could be consuming the
memory in this fashion, I can create a patch for our internal use.
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:[email protected]
