-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/14427/
-----------------------------------------------------------

Review request for qpid.


Bugs: QPID-5199
    https://issues.apache.org/jira/browse/QPID-5199


Repository: qpid


Description
-------

E.g. for management stats or for enforcing queue depth limits. Right now the 
broker only considers the content size for the 0-10 path (it uses the full 
transfer payload size for 1.0), whereas it would be nicer to include the header 
segment also. This is also more consistent with sizes as computed for AMQP 1.0 
messages.


Diffs
-----

  /trunk/qpid/cpp/src/qpid/broker/Exchange.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/Message.h 1528044 
  /trunk/qpid/cpp/src/qpid/broker/Message.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/Queue.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/QueueFlowLimit.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/ThresholdAlerts.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/amqp/Message.h 1528044 
  /trunk/qpid/cpp/src/qpid/broker/amqp/Message.cpp 1528044 
  /trunk/qpid/cpp/src/qpid/broker/amqp_0_10/MessageTransfer.h 1528044 
  /trunk/qpid/cpp/src/qpid/broker/amqp_0_10/MessageTransfer.cpp 1528044 
  /trunk/qpid/cpp/src/tests/MessageTest.cpp 1528044 
  /trunk/qpid/cpp/src/tests/QueueFlowLimitTest.cpp 1528044 
  /trunk/qpid/cpp/src/tests/QueuePolicyTest.cpp 1528044 
  /trunk/qpid/tests/src/py/qpid_tests/broker_0_10/stats.py 1528044 
  /trunk/qpid/tests/src/py/qpid_tests/broker_0_10/threshold.py 1528044 

Diff: https://reviews.apache.org/r/14427/diff/


Testing
-------

make test passes

The bulk of the patch is taken up in tweaking tests that make assumptions about 
the message sizes. The full size isn't always easy to predict (e.g. broker adds 
exchange in, client may add qpid.subject header in etc). These tests remain 
quite sensitive to specific implementations. No more so than they were, but 
they now additionally include an observed overhead for the headers.


Thanks,

Gordon Sim

Reply via email to