Well, if I keep the map of tasks=>Future, I can call cancel() on it, then in the async method I can check context.wasCancelCalled() and actually cancel it (probably throwing an exception). But then, if isDone() can't tell if the thread as ended or not, I have to find another way to notify my map back that the task is not running anymore.
Oh boy. [] Leo On Thu, Jan 14, 2016 at 3:29 PM, Romain Manni-Bucau <[email protected]> wrote: > depends if you want the control in the caller or in your async method > basically but isDone will have the constraints we talked about > > > 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> > > 2016-01-14 18:27 GMT+01:00 Leonardo K. Shikida <[email protected]>: > > > well, but I still have the option to keep a map os task=>Future and check > > if Future.isDone() right? > > > > [] > > > > Leo > > > > On Wed, Jan 13, 2016 at 4:24 AM, Romain Manni-Bucau < > [email protected] > > > > > wrote: > > > > > Well technically it should be ~possible adding in your impl a map of > > > task=>thread and getting the thread from the task to interrupt it but > > > results it would be quite hard to tune portably since you then need to > > tune > > > the async pool to support such an usage. > > > Le 13 janv. 2016 00:21, "Leonardo K. Shikida" <[email protected]> a > > écrit > > > : > > > > > > > I wish I could just kill -9 it :-D but I understand it's a java > > problem, > > > > not a tomee problem :-) > > > > > > > > I'll stick to SessionContext then > > > > > > > > thx > > > > > > > > [] > > > > > > > > Leo > > > > > > > > On Tue, Jan 12, 2016 at 8:54 PM, Romain Manni-Bucau < > > > [email protected] > > > > > > > > > wrote: > > > > > > > > > You dont like SessionContext solution? > > > > > Le 12 janv. 2016 23:52, "Leonardo K. Shikida" <[email protected]> > a > > > > écrit > > > > > : > > > > > > > > > > > right, but how can I reach them? It seems I'd need somehow to get > > > > > > AsynchronousCall > > > > > > from invoke() > > > > > > > > > > > > is there any "right way" to do that? > > > > > > > > > > > > [] > > > > > > > > > > > > Leo > > > > > > > > > > > > On Tue, Jan 12, 2016 at 7:44 PM, Romain Manni-Bucau < > > > > > [email protected] > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > think > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > https://github.com/apache/tomee/blob/120a33c7b4de07ae01c17978ea37d88a911ea860/container/openejb-core/src/main/java/org/apache/openejb/async/AsynchronousPool.java#L146 > > > > > > > should help ;) > > > > > > > > > > > > > > As thread.stop() is deprecated asynchronous tasks should check > > > their > > > > > > state > > > > > > > ("isRunning") through the context in a correct implementation > if > > > > theiy > > > > > > can > > > > > > > be cancelled. > > > > > > > > > > > > > > > > > > > > > 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> > > > > > > > > > > > > > > 2016-01-12 21:28 GMT+01:00 Leonardo K. Shikida < > > [email protected] > > > >: > > > > > > > > > > > > > > > actually, it seems that cancel() does not interrupt anything > in > > > > this > > > > > > > > context > > > > > > > > > > > > > > > > it just changes a flag for > > > > > > > > > > > > > > > > @Resource > > > > > > > > private SessionContext context; > > > > > > > > > > > > > > > > > > > > > > > > to context.wasCancelCalled() > > > > > > > > > > > > > > > > right? > > > > > > > > > > > > > > > > [] > > > > > > > > > > > > > > > > Leo > > > > > > > > > > > > > > > > On Tue, Jan 12, 2016 at 6:13 PM, Leonardo K. Shikida < > > > > > > [email protected]> > > > > > > > > wrote: > > > > > > > > > > > > > > > > > I understand the need to wrap the result in a AsyncResult > > > > > > > > > > > > > > > > > > My question is if inside the @Asynchronous method I have a > > > > > > > Thread.sleep() > > > > > > > > > and if the caller method keeps the Future object, if I call > > > > > > > > > Future.cancel(true), will it send an interrupt to the > > > > > @Asynchrnonous > > > > > > > > method > > > > > > > > > or will it be ignored? > > > > > > > > > > > > > > > > > > [] > > > > > > > > > > > > > > > > > > Leo > > > > > > > > > > > > > > > > > > On Tue, Jan 12, 2016 at 10:58 AM, Romain Manni-Bucau < > > > > > > > > > [email protected]> wrote: > > > > > > > > > > > > > > > > > >> the future returned to the caller can be cancelled, the > > future > > > > > > > instance > > > > > > > > >> you > > > > > > > > >> return can't since it is here just to match the returned > > type: > > > > > > > > >> > > > > > > > > >> public Future<Foo> asyncMethod() { > > > > > > > > >> return new Foo(); > > > > > > > > >> } > > > > > > > > >> > > > > > > > > >> this is really what you do but it doesnt compile so you > wrap > > > Foo > > > > > in > > > > > > an > > > > > > > > >> AsyncResult to match java typing but if you debug it is > not > > a > > > > > > > > AsyncResult > > > > > > > > >> that the caller get but a real Future. > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> 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> > > > > > > > > >> > > > > > > > > >> 2016-01-12 13:42 GMT+01:00 Leonardo K. Shikida < > > > > [email protected] > > > > > >: > > > > > > > > >> > > > > > > > > >> > Hi > > > > > > > > >> > > > > > > > > > >> > Is it possible to cancel a long-running @Asynchronous > > > method? > > > > > > > > >> > > > > > > > > > >> > My idea was to cancel the Future object, but according > to > > > > > > > > >> > > > > > > > > > >> > > > > > > http://tomee.apache.org/examples-trunk/async-methods/README.html > > > > > > > > >> > > > > > > > > > >> > "Important to note that the AsyncResult object the > > > > JobProcessor > > > > > > > > returns > > > > > > > > >> is > > > > > > > > >> > not the same Future object the caller is holding. It > would > > > > have > > > > > > been > > > > > > > > >> neat > > > > > > > > >> > if the real JobProcessor could just return String and > the > > > > > caller's > > > > > > > > >> version > > > > > > > > >> > of JobProcessor could return Future<String>, but we > didn't > > > see > > > > > any > > > > > > > way > > > > > > > > >> to > > > > > > > > >> > do that without adding more complexity. So the > AsyncResult > > > is > > > > a > > > > > > > simple > > > > > > > > >> > wrapper object. The container will pull the String out, > > > throw > > > > > the > > > > > > > > >> > AsyncResult away, then put the String in the *real* > Future > > > > that > > > > > > the > > > > > > > > >> caller > > > > > > > > >> > is holding." > > > > > > > > >> > > > > > > > > > >> > This thread also indicates that it's not defined in the > > EJB > > > > spec > > > > > > > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > http://stackoverflow.com/questions/16493381/cannot-cancel-asynchronous-call-to-ejb > > > > > > > > >> > > > > > > > > > >> > Any help is welcome. > > > > > > > > >> > > > > > > > > > >> > [] > > > > > > > > >> > > > > > > > > > >> > Leo > > > > > > > > >> > > > > > > > > > >> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
