-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/701/
-----------------------------------------------------------
(Updated 2011-05-10 21:25:11.588275)
Review request for qpid, Alan Conway, Gordon Sim, and Kim van der Riet.
Changes
-------
This refactor brings transient traffic performance back to pre-patch levels.
I've moved the whole flush-or-save logic into the clone() callback which is
invoked only if the message.transfer cannot be completed asynchronously. This
removes it from the transient fast-path.
Let me know what you think.
Summary
-------
Broker now issues an explicit flush for each received message that has the sync
flag set in the message.transfer command.
Broker also tracks all messages received that do _not_ have the sync flag set,
and will flush these messages on receipt of an execution.sync.
This addresses bug QPID-3252.
https://issues.apache.org/jira/browse/QPID-3252
Diffs (updated)
-----
/trunk/qpid/cpp/src/qpid/broker/SessionState.h 1101542
/trunk/qpid/cpp/src/qpid/broker/SessionState.cpp 1101542
Diff: https://reviews.apache.org/r/701/diff
Testing
-------
qpid unit & store unit
Thanks,
Kenneth
