On 01/04/2008, Gordon Sim <[EMAIL PROTECTED]> wrote: > Aidan Skinner wrote: > > On Tue, Apr 1, 2008 at 10:39 AM, Gordon Sim <[EMAIL PROTECTED]> wrote: > > > >> Aidan Skinner wrote: > >> > I'm happy to announce that M2.1 is now frozen and that beta 2 has been > >> > rolled and is available for testing from: > >> > http://people.apache.org/~aidan/qpid-M2.1-RC1 > >> > >> Whats the plan for languages other than java? Do we have an official > >> release manager for this release (and if so are they producing and > >> signing all the artifacts)? > > > > I'm the RM, mostly by dint of doing it. I haven't been signing the > > betas, although I should have signed the RC, sorry. I'll upload the > > other languages later (with signatures). > > > Fantastic; thanks Aidan! > Aidan,
These few issues should really make it in to the release. The QueueHousekeeping thread can bring down the broker https://issues.apache.org/jira/browse/QPID-886 A similar issue exists with the MC methods but hasn't been seen in practice... still. https://issues.apache.org/jira/browse/QPID-888 This issue is what highlighted it for me when testing with the BerkeleyDB MessageStore module. I would also like to address this issue for M2.1 https://issues.apache.org/jira/browse/QPID-889 None of these issues are a large change. The first two will add try{}finally{UNLOCK} blocks for the unlock code which is how it should have been done in the first place, it even says so in the ReentrantLock javadoc :) The last issue would be good to fix. The first ones will prevent the broker hanging the latter is a nice fix to prevent the broker barfing with the BerkeleyDB MessageStore. I will also try and add some tests that will catch any regression in this area. I will have the changes done today so you can rebuild the Java along with the other artifacts. Martin -- Martin Ritchie
