[ 
https://issues.apache.org/jira/browse/QPID-6361?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Keith Wall updated QPID-6361:
-----------------------------
    Description: 
Currently on the [0-8..0-91] path in the Java Broker, queue.declare with an 
argument of passive false will close the channel if the queue already exists 
but has different durability.   The message is "cannot re-declare queue qqqqq 
with different durability (was: xxxxx requested yyyyy)". 

As the default behaviour of the AMQP JMS Client 0-8..0-91 is to automatically 
declare queues as a side effect of consumer creation, this effectively means 
that the knowledge of a queue's required durability needs to be spread across 
all applications.  For installations typical of large institutions, this is 
operationally inconvenient and inflexible.

The relevant AMQP specifications make no statements regarding the required 
behaviour of an active queue.declare if the queue already exists with a 
different durability.

This change will remove the validation.  In future, if the queue already 
exists, the queue declare will succeed even if durability is different.  The 
validations around exclusivity,. auto-delete, and the non-passive flag are 
unaffected by this change.

  was:
Currently on the [0-8..0-91] path in the Java Broker, queue.declare with an 
argument of passive false will close the channel if the queue already exists 
but has different durability.   The message is "cannot re-declare queue qqqqq 
with different durability (was: xxxxx requested yyyyy)". 

As the default behaviour of the AMQP JMS Client 0-8..0-91 is to automatically 
declare queues as a side effect of consumer creation, this effectively means 
that the knowledge of a queue's required durability needs to be spread across 
all applications.  For installations typical of large installations, this is 
operationally inconvenient and inflexible.

The relevant AMQP specifications make no statements regarding the required 
behaviour of an active queue.declare if the queue already exists with a 
different durability.

This change will remove the validation.  In future, if the queue already 
exists, the queue declare will succeed even if durability is different.  The 
validations around exclusivity,. auto-delete, and the non-passive flag are 
unaffected by this change.


> [0-8..0-091] Active queue declare no longer to validate queue durability 
> -------------------------------------------------------------------------
>
>                 Key: QPID-6361
>                 URL: https://issues.apache.org/jira/browse/QPID-6361
>             Project: Qpid
>          Issue Type: Improvement
>          Components: Java Broker
>            Reporter: Keith Wall
>            Assignee: Keith Wall
>             Fix For: 0.31
>
>
> Currently on the [0-8..0-91] path in the Java Broker, queue.declare with an 
> argument of passive false will close the channel if the queue already exists 
> but has different durability.   The message is "cannot re-declare queue qqqqq 
> with different durability (was: xxxxx requested yyyyy)". 
> As the default behaviour of the AMQP JMS Client 0-8..0-91 is to automatically 
> declare queues as a side effect of consumer creation, this effectively means 
> that the knowledge of a queue's required durability needs to be spread across 
> all applications.  For installations typical of large institutions, this is 
> operationally inconvenient and inflexible.
> The relevant AMQP specifications make no statements regarding the required 
> behaviour of an active queue.declare if the queue already exists with a 
> different durability.
> This change will remove the validation.  In future, if the queue already 
> exists, the queue declare will succeed even if durability is different.  The 
> validations around exclusivity,. auto-delete, and the non-passive flag are 
> unaffected by this change.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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

Reply via email to