No, I really want the MDBs. I'm going to explore connecting the MDBs directly to the broker, bypassing the RA.
Setting maxSessions to 1 does work, the problem with that is that there is then one and only one instance of the MDB. I might ultimately have between 1000 and 10000 groups in a typical deployment, and the # of groups in a single deployment could change over time. I need one instance of the MDB per active group (likely not all the available groups will be active at any given instance in time). Hiram Chirino wrote: > > If you configure the maxSessions to be 1 on the ActivationSpec then it > should keep the grouping. > > But if that won't work for you, would the spring message containers be > a good enough substitute for MDBs? > > > On 1/12/07, JohnRobinson <[EMAIL PROTECTED]> wrote: >> >> Thanks James! I would say from the behavior I've been seeing that the RA >> is >> indeed ignoring the groupId. >> >> I created 4 consumers on the queue, deleted the outbound MDB, and moved >> the >> message handling logic into the new class. When I do that, the grouping >> is >> indeed respected and messages arrive at the SOAP endpoint in the intended >> order! Yay. That's the good news. >> >> The bad news is that we really want to use MDBs. Any idea when the RA >> might >> become group aware? >> >> -John >> >> >> James.Strachan wrote: >> > >> > >> > The type of transport doesn't matter at all - vm, tcp or whatever. I'm >> > confident things would work fine if you just created, say, 100 >> > MessageListeners (each with their own session so they can support >> > parallel consuming). >> > >> > I'd definitely try make a simpler test case - taking out points of >> > complication which could affect the order (like having multiple EJBs >> > or parallel soap requests which could all break order). Given a number >> > of messages on the queue, placed in order with the same JMSXGroupID - >> > we should just test if the RA dispatches them in order. I've a feeling >> > the RA may ignore the message groups contract and parallelize the >> > dispatching to concurrent MDBs. >> > >> > -- >> > >> > James >> > ------- >> > http://radio.weblogs.com/0112098/ >> > >> > >> >> -- >> View this message in context: >> http://www.nabble.com/Groups-tf2952450.html#a8302770 >> Sent from the ActiveMQ - User mailing list archive at Nabble.com. >> >> > > > -- > Regards, > Hiram > > Blog: http://hiramchirino.com > > -- View this message in context: http://www.nabble.com/Groups-tf2952450.html#a8371838 Sent from the ActiveMQ - User mailing list archive at Nabble.com.
