Description of problem: 1. When qpid-config user provides bad input when creating a queue/exchange, an exception is returned from the client, ie: qpid-tools-0.18: Failed: Exception: Exception from Agent: {u'error_code': 7L, u'error_text': 'illegal-argument: Cannot convert qpid.max_size to unsigned integer: F8:int64(-1) (qpid/broker/QueuePolicy.cpp:148)'} qpid-tools-0.14: Failed: SessionException: ExecutionException(error_code=531, command_id=serial(73), class_code=8, command_code=1, field_index=0, description=u'illegal-argument: Cannot convert qpid.max_size to unsigned integer: F8:int64(-1) (qpid/broker/QueuePolicy.cpp:143)', error_info={}, channel=1, id=serial(77)) ^^ Correct - appropriate exception is returned from both the versions, the queue/exchange is not created). on 2.2 packages (0.14) following error message is logged immediately to the broker log: 2012-11-14 16:15:27 error Execution exception: illegal-argument: Cannot convert qpid.max_size to unsigned integer: F8:int64(-1) (qpid/broker/QueuePolicy.cpp:143) But this does not occur on 0.18 packages. 2. Then the qpid-tools user corrects his bad input and use the same name for a queue/exchange. The queue/exchange is properly created, but after several seconds following message is logged to broker log: 2012-11-14 17:14:42 [Management] error Detected two management objects with the same identifier: 0-546-1--47(org.apache.qpid.broker:queue:q) (^^this does not occur on 0.14 packages) Maybe some management data do not get properly cleaned? I do not see any real impact of this error, management tools lists only the one properly created queue/exchange. So the only issue here at the moment is that the management error message is logged to broker log. Feel free to change the component to qpid-tools if you find it more appropriate. Version-Release number of selected component (if applicable): qpid-cpp-*-0.18-9 qpid-tools-0.18-6 How reproducible: 100% Steps to Reproduce: 1. qpid-config add queue q --max-queue-size=-1 2. qpid-config add queue q --max-queue-size=1 3. wait several seconds until following message is logged: 2012-11-14 17:14:42 [Management] error Detected two management objects with the same identifier: 0-546-1--47(org.apache.qpid.broker:queue:q) Actual results: Management error message in the broker log Expected results: No such message logged Additional info:
Chuck, please assess.
Yes this is a bug. Part 2 bug "error Detected two management objects with the same identifier" is a resource leak.
This issue has been fixed. Verified on rhel5.9, rhel6.4 Beta (i386 / x86_64) Packages used for testing: qpid-cpp-*-0.18-13 -> VERIFIED
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHSA-2013-0561.html