If there's an external event to the MDB like a timeout (which exception I cannot catch), then it should have something like a @Timeout annotation so I could customize how the event could be handled.
Not tomee's problem of course, tomee just implements a standard. :-( Gonna think on something else here. [] Leo On Wed, Apr 15, 2015 at 4:21 AM, Romain Manni-Bucau <[email protected]> wrote: > hmm, my proposal was to use UserTransaction more than that. About this > particular exemple we would need a maven project to reproduce it. > > > Romain Manni-Bucau > @rmannibucau <https://twitter.com/rmannibucau> | Blog > <http://rmannibucau.wordpress.com> | Github < > https://github.com/rmannibucau> | > LinkedIn <https://www.linkedin.com/in/rmannibucau> | Tomitriber > <http://www.tomitribe.com> > > 2015-04-15 4:05 GMT+02:00 Leonardo K. Shikida <[email protected]>: > > > Hi > > > > So I've changed my MDBs to stateless > > > > public class Worker implements MessageListener { > > > > @Resource(name="MyJmsConnectionFactory") > > private ConnectionFactory connectionFactory; > > > > @Resource(name="DriverJobQueue") > > private Queue queue; > > > > private Connection connection = null; > > private Session session = null; > > private MessageConsumer consumer = null; > > > > @PostConstruct > > public void initialize() { > > System.out.println("Instantiating "+this); > > try { > > connection = connectionFactory.createConnection(); > > connection.start(); //activemq site says it's here > > // Create a Session > > session = connection.createSession(false, > > Session.AUTO_ACKNOWLEDGE); > > > > // Create a MessageConsumer from the Session to the Topic or > > Queue > > consumer = session.createConsumer(this.queue); > > > > consumer.setMessageListener(this); > > > > // connection.start(); //some sites says it's here > > > > } catch(Exception e) { > > e.printStackTrace(); > > } finally { > > if (consumer != null) { > > try { > > consumer.close(); > > } catch (JMSException e) { > > e.printStackTrace(); > > } > > } > > if (session != null) { > > try { > > session.close(); > > } catch (JMSException e) { > > e.printStackTrace(); > > } > > } > > if (connection != null) { > > try { > > connection.close(); > > } catch (JMSException e) { > > e.printStackTrace(); > > } > > } > > } > > } > > > > @Override > > public void onMessage(Message msg) { > > System.out.println("On message for "+this); > > } > > > > openejb-jar.xml > > > > <?xml version="1.0"?> > > <openejb-jar xmlns="http://www.openejb.org/openejb-jar/1.1";> > > <ejb-deployment > > ejb-name="Worker" > > deployment-id="Worker" > > container-id="pseudo_mdb" /> > > </openejb-jar> > > > > tomee.xml > > > > <Container id="myAllContainer" type="STATELESS"> > > strictPooling = false > > </Container> > > > > <Container id="pseudo_mdb" type="STATELESS"> > > strictPooling = true > > maxSize = 5 > > minSize = 3 > > </Container> > > > > The container part seems to be right, since only the Worker bean is > > instantiated early. > > > > But for some reason, it's not consuming from the queue. > > > > Am I missing something here? > > > > Still, I feel I am running in circles here... because even if I have some > > control on the context, I still won't be able to catch the timeout > > exception :-( > > > > grrrrrrrrrrrr > > > > [] > > > > Leo > > > > On Tue, Apr 14, 2015 at 6:08 PM, Leonardo K. Shikida <[email protected]> > > wrote: > > > > > thx! > > > > > > [] > > > > > > Leo > > > > > > On Tue, Apr 14, 2015 at 6:03 PM, Romain Manni-Bucau < > > [email protected] > > > > wrote: > > > > > >> using openejb-jar.xml you can wire a container to some beans only. > > >> > > >> > > >> Romain Manni-Bucau > > >> @rmannibucau <https://twitter.com/rmannibucau> | Blog > > >> <http://rmannibucau.wordpress.com> | Github < > > >> https://github.com/rmannibucau> | > > >> LinkedIn <https://www.linkedin.com/in/rmannibucau> | Tomitriber > > >> <http://www.tomitribe.com> > > >> > > >> 2015-04-14 22:59 GMT+02:00 Leonardo K. Shikida <[email protected]>: > > >> > > >> > yeah, I think this will be the way to go > > >> > > > >> > since I can configure max/min instances for stateless ( > > >> > http://tomee.apache.org/statelesscontainer-config.html) I think I > can > > >> just > > >> > create a pool of stateless instances to work as MDBs that can deal > > with > > >> > timeout :-) > > >> > > > >> > is there any way to configure a container just for these specfic > > >> stateless > > >> > EJBs (so some other configuration can be used for stateless by > > default)? > > >> > > > >> > [] > > >> > > > >> > Leo > > >> > > > >> > On Tue, Apr 14, 2015 at 5:46 PM, Romain Manni-Bucau < > > >> [email protected] > > >> > > > > >> > wrote: > > >> > > > >> > > so just handle transactions yourself > > >> > > > > >> > > > > >> > > Romain Manni-Bucau > > >> > > @rmannibucau <https://twitter.com/rmannibucau> | Blog > > >> > > <http://rmannibucau.wordpress.com> | Github < > > >> > > https://github.com/rmannibucau> | > > >> > > LinkedIn <https://www.linkedin.com/in/rmannibucau> | Tomitriber > > >> > > <http://www.tomitribe.com> > > >> > > > > >> > > 2015-04-14 22:42 GMT+02:00 Lars-Fredrik Smedberg < > > [email protected] > > >> >: > > >> > > > > >> > > > Another option is to look at the redelivered property/redelivery > > >> count > > >> > > and > > >> > > > take action (write to db?) in case of redelivery.... requires > you > > >> to > > >> > > set a > > >> > > > redelivery policy / number of retries etc > > >> > > > > > >> > > > /LF > > >> > > > On Apr 14, 2015 10:39 PM, "Leonardo K. Shikida" < > > [email protected]> > > >> > > wrote: > > >> > > > > > >> > > > > Yup, but if the MDB timeout, I need to set something in the > DB, > > >> > that's > > >> > > > why > > >> > > > > I need to treat the timeout. > > >> > > > > > > >> > > > > If the job to be executed timeout, then it will probably > timeout > > >> for > > >> > > any > > >> > > > > retry, because it's caused by a poorly chosen parameter, so > > >> retries > > >> > are > > >> > > > not > > >> > > > > important here (although increasing timeouts on each retry > could > > >> be > > >> > an > > >> > > > > interesting option) > > >> > > > > > > >> > > > > [] > > >> > > > > > > >> > > > > Leo > > >> > > > > > > >> > > > > On Tue, Apr 14, 2015 at 5:27 PM, Romain Manni-Bucau < > > >> > > > [email protected] > > >> > > > > > > > >> > > > > wrote: > > >> > > > > > > >> > > > > > Hmm, with MDBs you have retries since they are > transactional. > > >> you > > >> > can > > >> > > > > even > > >> > > > > > configure the redelivery policy with an exponential backoff > > and > > >> so > > >> > on > > >> > > > if > > >> > > > > > really needed > > >> > > > > > > > >> > > > > > > > >> > > > > > Romain Manni-Bucau > > >> > > > > > @rmannibucau <https://twitter.com/rmannibucau> | Blog > > >> > > > > > <http://rmannibucau.wordpress.com> | Github < > > >> > > > > > https://github.com/rmannibucau> | > > >> > > > > > LinkedIn <https://www.linkedin.com/in/rmannibucau> | > > Tomitriber > > >> > > > > > <http://www.tomitribe.com> > > >> > > > > > > > >> > > > > > 2015-04-14 22:13 GMT+02:00 Leonardo K. Shikida < > > >> [email protected] > > >> > >: > > >> > > > > > > > >> > > > > > > Well, what I need is a pool of workers to listen to a > queue > > >> and > > >> > > > process > > >> > > > > > > messages while they come, but I also need to properly > treat > > >> them > > >> > > when > > >> > > > > the > > >> > > > > > > JMS transaction times out. > > >> > > > > > > > > >> > > > > > > So I guess I can't use a MDB here right? > > >> > > > > > > > > >> > > > > > > What would be a good approach in this situation? > > >> > > > > > > > > >> > > > > > > [] > > >> > > > > > > > > >> > > > > > > Leo > > >> > > > > > > > > >> > > > > > > On Tue, Apr 14, 2015 at 4:53 PM, Romain Manni-Bucau < > > >> > > > > > [email protected] > > >> > > > > > > > > > >> > > > > > > wrote: > > >> > > > > > > > > >> > > > > > > > if you want to control and be able to catch it you need > to > > >> > handle > > >> > > > > > > yourself > > >> > > > > > > > the transaction otherwise if it takes more then you'll > > get a > > >> > > > rollback > > >> > > > > > > and a > > >> > > > > > > > (surely wrapped) RollbackException > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > > > Romain Manni-Bucau > > >> > > > > > > > @rmannibucau <https://twitter.com/rmannibucau> | Blog > > >> > > > > > > > <http://rmannibucau.wordpress.com> | Github < > > >> > > > > > > > https://github.com/rmannibucau> | > > >> > > > > > > > LinkedIn <https://www.linkedin.com/in/rmannibucau> | > > >> > Tomitriber > > >> > > > > > > > <http://www.tomitribe.com> > > >> > > > > > > > > > >> > > > > > > > 2015-04-14 21:45 GMT+02:00 Leonardo K. Shikida < > > >> > > [email protected] > > >> > > > >: > > >> > > > > > > > > > >> > > > > > > > > just checking > > >> > > > > > > > > > > >> > > > > > > > > I just add > > >> > > > > > > > > > > >> > > > > > > > > @Resource > > >> > > > > > > > > private TransactionManager tx; > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > and in the MDB initialization I set some timeout > > >> > > > > > > > > > > >> > > > > > > > > @PostConstruct > > >> > > > > > > > > public void initialize() { > > >> > > > > > > > > try { > > >> > > > > > > > > tx.setTransactionTimeout(3); > > >> > > > > > > > > } catch (SystemException e) { > > >> > > > > > > > > e.printStackTrace(); > > >> > > > > > > > > } > > >> > > > > > > > > } > > >> > > > > > > > > > > >> > > > > > > > > and when my onMessage() executes, if it takes more > than > > 3 > > >> > > seconds > > >> > > > > to > > >> > > > > > > > > complete, it throws and exception? > > >> > > > > > > > > > > >> > > > > > > > > and if so, how do I catch it? (or I don't?) > > >> > > > > > > > > > > >> > > > > > > > > [] > > >> > > > > > > > > > > >> > > > > > > > > Leo > > >> > > > > > > > > > > >> > > > > > > > > On Tue, Apr 14, 2015 at 4:20 PM, Leonardo K. Shikida < > > >> > > > > > > [email protected]> > > >> > > > > > > > > wrote: > > >> > > > > > > > > > > >> > > > > > > > > > thx! > > >> > > > > > > > > > > > >> > > > > > > > > > [] > > >> > > > > > > > > > > > >> > > > > > > > > > Leo > > >> > > > > > > > > > > > >> > > > > > > > > > On Tue, Apr 14, 2015 at 4:16 PM, Romain Manni-Bucau > < > > >> > > > > > > > > [email protected] > > >> > > > > > > > > > > wrote: > > >> > > > > > > > > > > > >> > > > > > > > > >> Get the Transactionmanager injected (@Resource) and > > >> call > > >> > > > > > > > > >> setTransactionTimeout(seconds) > > >> > > > > > > > > >> > > >> > > > > > > > > >> > > >> > > > > > > > > >> Romain Manni-Bucau > > >> > > > > > > > > >> @rmannibucau <https://twitter.com/rmannibucau> | > > Blog > > >> > > > > > > > > >> <http://rmannibucau.wordpress.com> | Github < > > >> > > > > > > > > >> https://github.com/rmannibucau> | > > >> > > > > > > > > >> LinkedIn <https://www.linkedin.com/in/rmannibucau> > | > > >> > > > Tomitriber > > >> > > > > > > > > >> <http://www.tomitribe.com> > > >> > > > > > > > > >> > > >> > > > > > > > > >> 2015-04-14 19:38 GMT+02:00 Leonardo K. Shikida < > > >> > > > > [email protected] > > >> > > > > > >: > > >> > > > > > > > > >> > > >> > > > > > > > > >> > Hi > > >> > > > > > > > > >> > > > >> > > > > > > > > >> > How do I increase the MDB onMessage() transaction > > >> > timeout? > > >> > > > > > > > > >> > > > >> > > > > > > > > >> > [] > > >> > > > > > > > > >> > > > >> > > > > > > > > >> > Leo > > >> > > > > > > > > >> > > > >> > > > > > > > > >> > > >> > > > > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > >> > > > > > > > > >> > > > > > > > >> > > > > > > >> > > > > > >> > > > > >> > > > >> > > > > > > > > >
