[ 
https://issues.apache.org/jira/browse/QPID-4022?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Chuck Rolke updated QPID-4022:
------------------------------

    Attachment: QPID-4022-conn-limits-rev2-10-with-tests.patch

I put the last patch up for review a week ago but reviews are in Maintenance.

Here is the proposed patch, including fixes to the self tests.
                
> C++ Broker connection limits by host ip and by user name can get confused
> -------------------------------------------------------------------------
>
>                 Key: QPID-4022
>                 URL: https://issues.apache.org/jira/browse/QPID-4022
>             Project: Qpid
>          Issue Type: Bug
>          Components: C++ Broker
>    Affects Versions: 0.16
>            Reporter: Chuck Rolke
>            Assignee: Chuck Rolke
>         Attachments: QPID-4022-conn-limits-rev2-10-with-tests.patch
>
>
> The current ACL module uses the ConnectionObserver to watch the life cycle of 
> connections. It tries to disallow the creation of too many connections by a 
> user or from an IP address. However, the method is uses is flawed especially 
> in the cluster case.
> A better strategy to use it to provide approvers in the ConnectionObserver 
> scheme and then to call them:
> 1. Limits by IP address are disapproved in the ConnectionFactories. If the 
> limit is reached then the factory does not create the connection codec and 
> the connection never begins a life cycle. This is enforced at the same point 
> in code as the per-broker --max-connection limit using similar enforcement 
> methods.
> 2. Limits by user name are disapproved at the same point as user 
> authentication happens. Details to follow.

--
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

        

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to