[
https://issues.apache.org/jira/browse/QPID-3875?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13219422#comment-13219422
]
[email protected] commented on QPID-3875:
-----------------------------------------------------
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/4107/#review5463
-----------------------------------------------------------
Ship it!
(For consistency it might be nice to have the txn stats
(txnMsgEnqueues/Dequeues) use the same mechanism?)
- Gordon
On 2012-02-29 18:42:59, Kenneth Giusti wrote:
bq.
bq. -----------------------------------------------------------
bq. This is an automatically generated e-mail. To reply, visit:
bq. https://reviews.apache.org/r/4107/
bq. -----------------------------------------------------------
bq.
bq. (Updated 2012-02-29 18:42:59)
bq.
bq.
bq. Review request for qpid, Andrew Stitcher, Gordon Sim, and Ted Ross.
bq.
bq.
bq. Summary
bq. -------
bq.
bq. For each message received or sent, the broker needs to increment at least
4 statistic counters - # of received/send messages, and # of received/sent
bytes. For each counter accessed, a call to thread-local storage is made.
This patch reduces the number of calls to thread local storage by allowing
direct access to the full complement of per thread counters.
bq.
bq. On my cpu-challenged laptop, this resulted in a minor speedup in msgs/sec
across a shared queue (2 senders, 1 receiver):
bq.
bq. TRUNK
bq. msg/sec
bq. Sender1 : Sender2 : Receiver
bq. 25825 : 25359 : 43805
bq. 25251 : 25159 : 43379
bq. 24564 : 24211 : 38836
bq.
bq. PATCHED
bq. msg/sec
bq. Sender1 : Sender2 : Receiver
bq. 26015 : 25849 : 44473
bq. 26197 : 25765 : 44291
bq. 25826 : 25476 : 43904
bq.
bq.
bq. This addresses bug qpid-3875.
bq. https://issues.apache.org/jira/browse/qpid-3875
bq.
bq.
bq. Diffs
bq. -----
bq.
bq. /trunk/qpid/cpp/src/qpid/broker/Exchange.cpp 1295123
bq. /trunk/qpid/cpp/src/qpid/broker/Queue.h 1295123
bq. /trunk/qpid/cpp/src/qpid/broker/Queue.cpp 1295123
bq. /trunk/qpid/cpp/managementgen/qmfgen/templates/Class.h 1295123
bq. /trunk/qpid/cpp/src/qpid/broker/Connection.cpp 1295123
bq.
bq. Diff: https://reviews.apache.org/r/4107/diff
bq.
bq.
bq. Testing
bq. -------
bq.
bq. make check
bq.
bq.
bq. Thanks,
bq.
bq. Kenneth
bq.
bq.
> Reduce processing overhead when updating multiple statistic counters
> --------------------------------------------------------------------
>
> Key: QPID-3875
> URL: https://issues.apache.org/jira/browse/QPID-3875
> Project: Qpid
> Issue Type: Improvement
> Components: C++ Broker
> Affects Versions: 0.16
> Reporter: Ken Giusti
> Assignee: Ken Giusti
> Priority: Trivial
> Fix For: Future
>
> Attachments: CG-statspatch.tgz
>
>
> While running cachegrind analysis of a simple traffic flow across the broker,
> I noticed that fetching the per-thread statistics had an overall instruction
> cost that is surprisingly long.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:[email protected]
