BTW, I assume you're talking about ActiveMQ Artemis since you're referencing QueueManagerImpl and ExpiryScanner (which are both classes in the Artemis code-base but not in the 5.x code-base).
Justin On Fri, Nov 9, 2018 at 9:54 AM Justin Bertram <jbert...@apache.org> wrote: > I believe you can get the behavior you want in 2.7.0 via ARTEMIS-1856 > [1]. You'll just need to configure a non-zero delay so that the broker > schedules the AddressQueueReaper. > > > Justin > > [1] https://issues.apache.org/jira/browse/ARTEMIS-1856 > > On Fri, Nov 9, 2018 at 9:45 AM Tim Bain <tb...@alumni.duke.edu> wrote: > >> When the messages expire, are they actually deleted from the broker? By >> default, there is nothing that will delete the expired messages in the >> absence of a consumer (or queue browser), so it might be that the queue >> isn't being deleted because it isn't actually empty. >> >> If that is indeed what's going on and you'd like ActiveMQ to behave >> differently (probably to have the broker periodically check for expired >> messages and delete them, even if there's no consumer), you can submit an >> enhancement request in JIRA for it. >> >> Tim >> >> On Thu, Nov 8, 2018, 11:23 AM pholik1 <petr_ho...@kb.cz wrote: >> >> > Hello, >> > >> > when there is autocreated queue with messages and theese messages >> expires, >> > the queue is left abandoned on server until somebody connects to it and >> > disconects - it triggers QueueManagerImpl.run() which destroys the >> queue. >> > In >> > my opinion there should be some code for check if queue is eligble for >> > autodelete in ExpiryScanner. >> > >> > Or is this behavior expected? >> > >> > Sincerely Petr Holik >> > >> > >> > >> > -- >> > Sent from: >> > http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html >> > >> >
