----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13650/ -----------------------------------------------------------
Review request for qpid and Chug Rolke. Bugs: https://bugzilla.redhat.com/show_bug.cgi?id=955688 https://issues.apache.org/jira/browse/https://bugzilla.redhat.com/show_bug.cgi?id=955688 Repository: qpid Description ------- When creating/deleting durable queues, their "owner" isn't tracked after restart. Therefore it is possible to exceed queue quotas after a restart. Problem scenario 1: -Set ACL limit on a user to 10 queues -Create 10 durable queues for that user -Restart the broker -The 10 durable queues are automatically recovered -The user can now create an additional 10 queues Problem scenario 2: - Set ACL limit on a user to 10 queues - Create 10 durable queues for that user - Decrease ACL limit on that user to 5 queues - Restart the broker - All 10 queues should be recovered, but no new ones should be allow until user drops below 5 queues This is a patch supplied by Pavel Moravec. It allows all existing durable queues to be recovered during broker restart, but prevents any new queues from being created until the user is below the ACL limit. Diffs ----- /trunk/qpid/cpp/src/qpid/broker/Broker.cpp 1515496 /trunk/qpid/cpp/src/qpid/broker/Queue.h 1515496 /trunk/qpid/cpp/src/qpid/broker/Queue.cpp 1515496 /trunk/qpid/cpp/src/qpid/broker/QueueRegistry.cpp 1515496 Diff: https://reviews.apache.org/r/13650/diff/ Testing ------- Tried patch with both scenarios from description. Thanks, Ernie Allen
