You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@qpid.apache.org by "Ted Ross (JIRA)" <qp...@incubator.apache.org> on 2008/05/22 15:41:56 UTC
[jira] Created: (QPID-1088) Locking cleanup for management objects
Locking cleanup for management objects
--------------------------------------
Key: QPID-1088
URL: https://issues.apache.org/jira/browse/QPID-1088
Project: Qpid
Issue Type: Improvement
Components: C++ Broker
Affects Versions: M3
Reporter: Ted Ross
Assignee: Ted Ross
Priority: Minor
Fix For: M3
A recent change to the management code generator caused excessive locking during the increment and decrement of counter statistics in management objects.
This update does the following:
1) Change RW locks in management objects to Mutexes
2) Remove locking for increment and decrement operations on counters
3) Provide user access to the management object mutex
4) Add higher-level locking in the broker::Queue class to protect all counter operations as a block.
There should now be only one lock taken out for management during the production or consumption of a message.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (QPID-1088) Locking cleanup for management objects
Posted by "Ted Ross (JIRA)" <qp...@incubator.apache.org>.
[ https://issues.apache.org/jira/browse/QPID-1088?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ted Ross resolved QPID-1088.
----------------------------
Resolution: Fixed
> Locking cleanup for management objects
> --------------------------------------
>
> Key: QPID-1088
> URL: https://issues.apache.org/jira/browse/QPID-1088
> Project: Qpid
> Issue Type: Improvement
> Components: C++ Broker
> Affects Versions: M3
> Reporter: Ted Ross
> Assignee: Ted Ross
> Priority: Minor
> Fix For: M3
>
>
> A recent change to the management code generator caused excessive locking during the increment and decrement of counter statistics in management objects.
> This update does the following:
> 1) Change RW locks in management objects to Mutexes
> 2) Remove locking for increment and decrement operations on counters
> 3) Provide user access to the management object mutex
> 4) Add higher-level locking in the broker::Queue class to protect all counter operations as a block.
> There should now be only one lock taken out for management during the production or consumption of a message.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.