qpid-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chug Rolke" <cro...@redhat.com>
Subject Re: Review Request: C++ Broker queue creation limits require better granularity
Date Fri, 01 Mar 2013 19:47:39 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/9703/#review17258
-----------------------------------------------------------


This is a promising first cut. Thanks for posting the patch.


trunk/qpid/cpp/src/qpid/acl/AclPlugin.cpp
<https://reviews.apache.org/r/9703/#comment36690>

    Don't put a queue count here as this forces a default into the command line switch. When
that happens then quotas can never be disabled by have no cli setting and no rules in the
rule file.



trunk/qpid/cpp/src/qpid/acl/AclResourceCounter.cpp
<https://reviews.apache.org/r/9703/#comment36689>

    The comparison for queue resources should be "<" not "<=". For connections the same
line is <= but connections have a different life cycle.



trunk/qpid/cpp/src/tests/acl.py
<https://reviews.apache.org/r/9703/#comment36695>

    I'd like to see the name of the queue to be test-globally unique. The queue declare may
fail because of a quota limit or becuse the queue already exists. If you use unique names
then you can isolate the creation error to a quota denial.



trunk/qpid/cpp/src/tests/acl.py
<https://reviews.apache.org/r/9703/#comment36691>

    when a queue declare fails they you get the exception but as a side effect the session
is gone. In order for the test to keep going you have to recreate a new session. See test_queue_allow_mode
for example that does this. 



trunk/qpid/cpp/src/tests/acl.py
<https://reviews.apache.org/r/9703/#comment36692>

    probably need a try/catch here


- Chug Rolke


On March 1, 2013, 5:07 p.m., Ernie Allen wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/9703/
> -----------------------------------------------------------
> 
> (Updated March 1, 2013, 5:07 p.m.)
> 
> 
> Review request for qpid and Chug Rolke.
> 
> 
> Description
> -------
> 
> Implements acl queue limits per user.
> - Modifies existing code to parse the acl file
> - Adds code to count queues per user
> - Adds max queue count of 65530
> - Adds python tests
> 
> 
> This addresses bug QPID-4604.
>     https://issues.apache.org/jira/browse/QPID-4604
> 
> 
> Diffs
> -----
> 
>   trunk/qpid/cpp/src/qpid/acl/Acl.cpp 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclData.h 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclData.cpp 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclPlugin.cpp 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclReader.h 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclReader.cpp 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclResourceCounter.h 1450922 
>   trunk/qpid/cpp/src/qpid/acl/AclResourceCounter.cpp 1450922 
>   trunk/qpid/cpp/src/tests/acl.py 1451307 
> 
> Diff: https://reviews.apache.org/r/9703/diff/
> 
> 
> Testing
> -------
> 
> Passes python tests:
> - limit on a named user
> - limit on a user in a group
> - two acl entries for same user, last entry is honored
> - limit of 0 on user is enforced
> - if all is specified, unnamed user gets all's limit
> - if all is not specified, unnamed user is denied
> 
> 
> Thanks,
> 
> Ernie Allen
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message