At first glance a single lock should work - removing the sm lock should prevent deadlocks down the road.
I'm not sure if I can reproduce the problem on my laptop - might have to wait until I get home to try it out. Thanks for the pointer to OpenJPA-453 - lots of data there I'll have to read up a bit. -mike On Mon, Aug 16, 2010 at 1:29 PM, Ravi Palacherla <[email protected] > wrote: > This patch does broker lock and in that case, is there a need for sm lock > too ? > Can we eliminate the sm lock and rely completely on broker lock? > The patch for the above is in 453. > > Regards, > Ravi. > > -----Original Message----- > From: Rick Curtis [mailto:[email protected]] > Sent: Monday, August 16, 2010 12:25 PM > To: [email protected] > Subject: Re: svn commit: r985067 - > /openjpa/trunk/openjpa-slice/src/test/java/org/apache/openjpa/slice/TestQueryMultiThreaded.java > > Go ahead and pull the trigger. > > On Mon, Aug 16, 2010 at 1:20 PM, Michael Dick <[email protected] > >wrote: > > > I understand your concern about giving people a false sense of confidence > - > > openjpa.Multithreaded requires more than just this fix. OTOH it's going > to > > be a while before we have the whole solution. > > > > If you don't have a specific concern with the patch I'll go ahead and > > commit > > it - I don't see it as dangerous for normal (non openpja.Multithreaded) > use > > and we have at least one testcase which depends on it. > > > > -mike > > > > On Mon, Aug 16, 2010 at 8:50 AM, Rick Curtis <[email protected]> wrote: > > > > > > > > My biggest concern with that fix is that it isn't a complete fix for > the > > > openjpa.Multithreaded property. It fixes the one one reported bug, but > > > there > > > are numerous others out there. > > > > > > I'd really like to see someone find some time to come up with a > complete > > > multi threaded story for openjpa. > > > > >
