[ 
https://issues.apache.org/jira/browse/QPIDJMS-464?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Robbie Gemmell updated QPIDJMS-464:
-----------------------------------
    Description: 
When a receive with timeout call is made and the client configured to use zero 
prefetch, the client issues a credit, waits for a message to arrive, and if one 
doesn't in the given time it drains the credit to either get a message or 
ensure there is none available and remove the credit. If the message is already 
partially transferred (i.e it is multi-framed and so likely a larger message) 
at this point the client knows it doesn't need to drain, but the way it handles 
this triggers a later mishandling of completing the receive request when the 
message arrives.

Some related comments can be found 
[here|https://issues.apache.org/jira/browse/QPIDJMS-458?focusedCommentId=16872498&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16872498]
 onwards.

  was:
When a receive with timeout call is made using zero prefetch, the client issues 
a credit, wait for a message to arrive, and if one doesn't in the given time it 
drains the credit to either get a message or ensure there is none and remove 
the credit. If the message is already partially transfered (i.e it is likely 
larger, and typically multi-framed though it need not be) at this point the 
client knows it doesnt need to drain, but the way it handles this triggers a 
later mishandling of completing the receive request when the message arrives.

Some related comments can be found 
[here|https://issues.apache.org/jira/browse/QPIDJMS-458?focusedCommentId=16872498&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16872498]
 onwards.


> zero prefetch receive can hang if timeout reached whilst message is partially 
> transferred
> -----------------------------------------------------------------------------------------
>
>                 Key: QPIDJMS-464
>                 URL: https://issues.apache.org/jira/browse/QPIDJMS-464
>             Project: Qpid JMS
>          Issue Type: Bug
>          Components: qpid-jms-client
>    Affects Versions: 0.43.0
>            Reporter: Robbie Gemmell
>            Assignee: Robbie Gemmell
>            Priority: Major
>             Fix For: 0.44.0
>
>
> When a receive with timeout call is made and the client configured to use 
> zero prefetch, the client issues a credit, waits for a message to arrive, and 
> if one doesn't in the given time it drains the credit to either get a message 
> or ensure there is none available and remove the credit. If the message is 
> already partially transferred (i.e it is multi-framed and so likely a larger 
> message) at this point the client knows it doesn't need to drain, but the way 
> it handles this triggers a later mishandling of completing the receive 
> request when the message arrives.
> Some related comments can be found 
> [here|https://issues.apache.org/jira/browse/QPIDJMS-458?focusedCommentId=16872498&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16872498]
>  onwards.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

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

Reply via email to