Bug 1064937
| Summary: | x-declare options auto-delete and exclusive to be ignored for topic subscriptions | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise MRG | Reporter: | Pavel Moravec <pmoravec> | ||||
| Component: | python-qpid | Assignee: | messaging-bugs <messaging-bugs> | ||||
| Status: | CLOSED NOTABUG | QA Contact: | MRG Quality Engineering <mrgqe-bugs> | ||||
| Severity: | low | Docs Contact: | |||||
| Priority: | low | ||||||
| Version: | 3.0 | CC: | gsim | ||||
| Target Milestone: | --- | Keywords: | Patch, TestCaseProvided | ||||
| Target Release: | --- | ||||||
| Hardware: | All | ||||||
| OS: | All | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2014-02-17 07:28:30 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
Created attachment 862842 [details]
Trivial patch
Remove "auto-delete" and "exclusive" keys from declare dictionary when creating the queue for topic subscription.
Also, it makes sense to simplify:
auto_delete=(reliability == "unreliable")
to:
auto_delete=True
as it must be set everytime (and already is, see how is reliability set few lines above).
My concern is that changing this breaks backward compatibility, which I'm not convinced is worth it. (In reply to Gordon Sim from comment #2) > My concern is that changing this breaks backward compatibility, which I'm > not convinced is worth it. C++ and Java clients ignore the flags, while Python client does not. That was my concern of raising the BZ, to unify clients behaviour. Is there a scenario when it makes sense to suppress auto-delete or exclusive flags (just) in Python client for topic subscription? (In reply to Pavel Moravec from comment #3) > C++ and Java clients ignore the flags, while Python client does not. That > was my concern of raising the BZ, to unify clients behaviour. Understood, and that is a good thing in general. However if an application currently is relying on setting them explicitly in python, taking that ability away would break their system. Also, I'm pretty sure the c++ client doesn't ignore the exclusive flag. It does ignore the auto-delete flag at present, inferring the value based on durability or reliability of the link. But perhaps the better fix is to change the c++ client (and java) to recognise these flags if explicitly specified. > Is there a scenario when it makes sense to suppress auto-delete or exclusive > flags (just) in Python client for topic subscription? (In reply to Gordon Sim from comment #4) > (In reply to Pavel Moravec from comment #3) > > C++ and Java clients ignore the flags, while Python client does not. That > > was my concern of raising the BZ, to unify clients behaviour. > > Understood, and that is a good thing in general. However if an application > currently is relying on setting them explicitly in python, taking that > ability away would break their system. > > Also, I'm pretty sure the c++ client doesn't ignore the exclusive flag. It > does ignore the auto-delete flag at present, inferring the value based on > durability or reliability of the link. But perhaps the better fix is to > change the c++ client (and java) to recognise these flags if explicitly > specified. > It makes sense to have the ability in Java and C++ clients, indeed. So I will raise BZs for that. P.S.: Currently, C++ client ignores auto-delete flag, while Java client ignores both flags. *** Bug 1063323 has been marked as a duplicate of this bug. *** |
Description of problem: Creating a topic subscription with x-declare:{'auto-delete':False, 'exclusive':False} in Python client sets the auxiliary queue as NON-auto-delete and NON-exclusive. This differs from C++ as well as Java client and is IMHO wrong: the auxiliary queue has to be auto-delete and exclusive every time and the two x-declare options should be ignored. Version-Release number of selected component (if applicable): python-qpid-0.22-11.el6.noarch How reproducible: 100% Steps to Reproduce: python qpid-receive -a "amq.direct/#; {link: {x-declare:{'auto-delete':False, 'exclusive':False}}}" -m 1 --print-content=no -f & python qpid-receive -a "amq.direct/#; {link: {x-declare:{'auto-delete':False}}}" -m 1 --print-content=no -f & python qpid-receive -a "amq.direct/#; {link: {x-declare:{'exclusive':False}}}" -m 1 --print-content=no -f & qpid-stat -q Actual results: qpid-stat -q shows one queue with no autoDel or excl, one with no autoDel and one with no excl flags. (plus one queue with both flags set - that is the queue for qpid-stat itself). Expected results: All queues to have both autoDel and excl flags set. Additional info: