You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@qpid.apache.org by Kenneth Giusti <kg...@apache.org> on 2011/08/31 22:37:49 UTC
Review Request: QPID-3346: add functional test tool for verifying message
group behavior.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1690/
-----------------------------------------------------------
Review request for qpid.
Summary
-------
Update to message group development:
Created a functional test tool that verifies message group behavior across multiple clients consuming from a single queue.
Also, bugfix: the "tag" used to identify a Consumer is not unique beyond the consumer's session. Queue's need to track consumers across different sessions, and using the tag was not unique enough. I've introduced a consumer name that is generated on the broker and should be unique broker-wide.
This addresses bug qpid-3346.
https://issues.apache.org/jira/browse/qpid-3346
Diffs
-----
/branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h 1158073
/branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.h 1158073
/branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.cpp 1158073
/branches/qpid-3346/qpid/cpp/src/qpid/cluster/UpdateClient.cpp 1158073
/branches/qpid-3346/qpid/cpp/src/tests/Makefile.am 1158073
/branches/qpid-3346/qpid/cpp/src/tests/msg_group_test.cpp PRE-CREATION
/branches/qpid-3346/qpid/cpp/src/tests/run_msg_group_tests PRE-CREATION
Diff: https://reviews.apache.org/r/1690/diff
Testing
-------
One simple functional test as an example - will flesh out the test script with multiple runs of the tool with different consumer and producer settings.
Thanks,
Kenneth
Re: Review Request: QPID-3346: add functional test tool for verifying
message group behavior.
Posted by Alan Conway <ac...@redhat.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1690/#review1721
-----------------------------------------------------------
/branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h
<https://reviews.apache.org/r/1690/#comment3927>
Why change the meaning of an existing member rather than introduce a new member like id?
- Alan
On 2011-08-31 20:37:49, Kenneth Giusti wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/1690/
> -----------------------------------------------------------
>
> (Updated 2011-08-31 20:37:49)
>
>
> Review request for qpid.
>
>
> Summary
> -------
>
> Update to message group development:
>
> Created a functional test tool that verifies message group behavior across multiple clients consuming from a single queue.
>
> Also, bugfix: the "tag" used to identify a Consumer is not unique beyond the consumer's session. Queue's need to track consumers across different sessions, and using the tag was not unique enough. I've introduced a consumer name that is generated on the broker and should be unique broker-wide.
>
>
> This addresses bug qpid-3346.
> https://issues.apache.org/jira/browse/qpid-3346
>
>
> Diffs
> -----
>
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/cluster/UpdateClient.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/Makefile.am 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/msg_group_test.cpp PRE-CREATION
> /branches/qpid-3346/qpid/cpp/src/tests/run_msg_group_tests PRE-CREATION
>
> Diff: https://reviews.apache.org/r/1690/diff
>
>
> Testing
> -------
>
> One simple functional test as an example - will flesh out the test script with multiple runs of the tool with different consumer and producer settings.
>
>
> Thanks,
>
> Kenneth
>
>
Re: Review Request: QPID-3346: add functional test tool for verifying
message group behavior.
Posted by Gordon Sim <gs...@redhat.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1690/#review1722
-----------------------------------------------------------
/branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h
<https://reviews.apache.org/r/1690/#comment3932>
Perhaps we could simply revert to using the address of the consumer pointer?
- Gordon
On 2011-08-31 20:37:49, Kenneth Giusti wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/1690/
> -----------------------------------------------------------
>
> (Updated 2011-08-31 20:37:49)
>
>
> Review request for qpid.
>
>
> Summary
> -------
>
> Update to message group development:
>
> Created a functional test tool that verifies message group behavior across multiple clients consuming from a single queue.
>
> Also, bugfix: the "tag" used to identify a Consumer is not unique beyond the consumer's session. Queue's need to track consumers across different sessions, and using the tag was not unique enough. I've introduced a consumer name that is generated on the broker and should be unique broker-wide.
>
>
> This addresses bug qpid-3346.
> https://issues.apache.org/jira/browse/qpid-3346
>
>
> Diffs
> -----
>
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/cluster/UpdateClient.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/Makefile.am 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/msg_group_test.cpp PRE-CREATION
> /branches/qpid-3346/qpid/cpp/src/tests/run_msg_group_tests PRE-CREATION
>
> Diff: https://reviews.apache.org/r/1690/diff
>
>
> Testing
> -------
>
> One simple functional test as an example - will flesh out the test script with multiple runs of the tool with different consumer and producer settings.
>
>
> Thanks,
>
> Kenneth
>
>
Re: Review Request: QPID-3346: add functional test tool for verifying
message group behavior.
Posted by Kenneth Giusti <kg...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1690/#review1725
-----------------------------------------------------------
/branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h
<https://reviews.apache.org/r/1690/#comment3936>
Possibly - if the order of consumers in my container doesn't matter: think state replication across a cluster (I got burned by that exact issue before - different addresses caused ordering inconsistencies).
I think order doesn't matter - I'll give this a try as it would clean things up a bit.
/branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h
<https://reviews.apache.org/r/1690/#comment3937>
Oops - actually, this diff doesn't show that I -introduced- this member on this branch awhile back. Originally, I removed the private "tag" from ConsumerImpl and replaced it with the "name" in the base class (so clients could be referred to by name).
Per Gordon's comment, I'm going to try to revert these consumer changes (restore the trunk version) if possible. Stay tuned....
- Kenneth
On 2011-08-31 20:37:49, Kenneth Giusti wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/1690/
> -----------------------------------------------------------
>
> (Updated 2011-08-31 20:37:49)
>
>
> Review request for qpid.
>
>
> Summary
> -------
>
> Update to message group development:
>
> Created a functional test tool that verifies message group behavior across multiple clients consuming from a single queue.
>
> Also, bugfix: the "tag" used to identify a Consumer is not unique beyond the consumer's session. Queue's need to track consumers across different sessions, and using the tag was not unique enough. I've introduced a consumer name that is generated on the broker and should be unique broker-wide.
>
>
> This addresses bug qpid-3346.
> https://issues.apache.org/jira/browse/qpid-3346
>
>
> Diffs
> -----
>
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/Consumer.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.h 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/broker/SemanticState.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/qpid/cluster/UpdateClient.cpp 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/Makefile.am 1158073
> /branches/qpid-3346/qpid/cpp/src/tests/msg_group_test.cpp PRE-CREATION
> /branches/qpid-3346/qpid/cpp/src/tests/run_msg_group_tests PRE-CREATION
>
> Diff: https://reviews.apache.org/r/1690/diff
>
>
> Testing
> -------
>
> One simple functional test as an example - will flesh out the test script with multiple runs of the tool with different consumer and producer settings.
>
>
> Thanks,
>
> Kenneth
>
>