qpid-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chug Rolke" <cro...@redhat.com>
Subject Re: Review Request 14000: QPID-5107 Trace queue&session deletion statistics show zero values for some counters everytime
Date Fri, 06 Sep 2013 13:41:48 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/14000/#review25959
-----------------------------------------------------------

Ship it!


This change arises from real-world scrutiny of the original print-stats-at-object-deletion
strategy. 

- Chug Rolke


On Sept. 6, 2013, 12:56 p.m., Pavel Moravec wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/14000/
> -----------------------------------------------------------
> 
> (Updated Sept. 6, 2013, 12:56 p.m.)
> 
> 
> Review request for qpid and Chug Rolke.
> 
> 
> Bugs: https://issues.apache.org/jira/browse/QPID-5107
>     https://issues.apache.org/jira/browse/https://issues.apache.org/jira/browse/QPID-5107
> 
> 
> Repository: qpid
> 
> 
> Description
> -------
> 
> qpid trace/logs statistics about object deletion. However some of these data are wrong,
if deletion process updates some stats. In particular msgDepth for a queue is everytime zero
(and msgTotalDequeues equals to msgTotalEnqueues despite no consumer was subscribed to the
queue), or unackedMessages for a session is zero everytime as well.
> 
> See https://issues.apache.org/jira/browse/https://issues.apache.org/jira/browse/QPID-5107
for two reproducer scenarios (deleting queue and closing session).
> 
> The problem is, that the stats are printed in destructor of the QMF object - too late.
I propose:
> - moving the code to debugStats() method
> - call the method from every manageable object at the beginning of its delete/destructor
method
> 
> 
> Diffs
> -----
> 
>   /trunk/qpid/cpp/managementgen/qmfgen/templates/Class.h 1520240 
>   /trunk/qpid/cpp/managementgen/qmfgen/templates/Class.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/Broker.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/DirectExchange.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/Exchange.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/FanOutExchange.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/HeadersExchange.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/Queue.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/SemanticState.cpp 1520245 
>   /trunk/qpid/cpp/src/qpid/broker/SessionState.cpp 1520245 
>   /trunk/qpid/cpp/src/qpid/broker/System.h 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/System.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/TopicExchange.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/Vhost.h 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/Vhost.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/broker/amqp_0_10/Connection.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/legacystore/MessageStoreImpl.cpp 1520240 
>   /trunk/qpid/cpp/src/qpid/xml/XmlExchange.cpp 1520240 
> 
> Diff: https://reviews.apache.org/r/14000/diff/
> 
> 
> Testing
> -------
> 
> Successfully tried the patch with both scenarios.
> 
> 
> Thanks,
> 
> Pavel Moravec
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message