Is it reproducible? Seems that some timers do not get canceled...
On Sat, Oct 11, 2014 at 12:07 PM, Robert Metzger <[email protected]> wrote: > WordCount [Java Application] > org.apache.flink.examples.java.wordcount.WordCount at localhost:38214 > (Suspended) > Daemon System Thread [Signal Dispatcher] (Suspended) > Daemon System Thread [Finalizer] (Suspended) > Daemon System Thread [Reference Handler] (Suspended) > Thread [Timer-1] (Suspended) > waiting for: TaskQueue (id=193) > Object.wait(long) line: not available [native method] > TimerThread.mainLoop() line: 552 > TimerThread.run() line: 505 > Thread [Timer-3] (Suspended) > waiting for: TaskQueue (id=192) > Object.wait(long) line: not available [native method] > TimerThread.mainLoop() line: 552 > TimerThread.run() line: 505 > Thread [DestroyJavaVM] (Suspended) > /opt/java/bin/java (Oct 11, 2014, 12:02:52 PM) > > > > On Fri, Oct 10, 2014 at 7:31 PM, Stephan Ewen <[email protected]> wrote: > > > Which threads are active when the shutdown hangs? > > Am 10.10.2014 19:25 schrieb "Robert Metzger" <[email protected]>: > > > > > Ah damn .. I meant the "Scala API Documentation" ;) > > > > > > I just found an issue: If I'm running an example from eclipse, the JVM > > > process won't terminate properly after the job has finished. > > > It seems that a "Destroy JVM" thread is started and the main() thread > has > > > already gone away, but still, the JVM is not stopping. > > > > > > > > > On Fri, Oct 10, 2014 at 6:53 PM, Aljoscha Krettek <[email protected] > > > > > wrote: > > > > > > > Yes, the Scala API should be up to date. The latest master also has > > > > all the Operator IT Cases from Java ported to Java. They also contain > > > > POJOs and nested key selectors. > > > > > > > > On Fri, Oct 10, 2014 at 6:32 PM, Robert Metzger <[email protected] > > > > > > wrote: > > > > > Thank you Leonidas for helping with the test. > > > > > > > > > > I've successfully used the maven archetypes and pulled > 0.7-incubating > > > > from > > > > > the staged repository, so that seems to work. > > > > > I found and fixed a bug with the YARN deployment and tested it on > > > Hadoop > > > > > 2.4.0 using Amazon EMR (8 nodes). > > > > > I also tested Flink on a 24 nodes cluster with K-Means and > different > > > > > flavors of Wordcount... All good (except for this issue: > > > > > https://issues.apache.org/jira/browse/FLINK-1152 ) > > > > > > > > > > Has the Scala API been updated to reflect the newly integrated POJO > > > > support? > > > > > > > > > > Could somebody who's using Windows try it there? > > > > > > > > > > We should check if the web-client shows the optimizer plans of the > > > > examples > > > > > properly and if they all run without data. > > > > > > > > > > > > > > > > > > > > > > > > > On Wed, Oct 8, 2014 at 8:50 PM, Robert Metzger < > [email protected]> > > > > wrote: > > > > > > > > > >> No, its just a dry run and I want to do some tests on our cluster > > > first. > > > > >> > > > > >> Last time, I created a lot of release candidates and vote threads. > > > This > > > > >> time, I wanted to be smart and do the testing first ;) > > > > >> > > > > >> > > > > >> > > > > >> On Wed, Oct 8, 2014 at 8:46 PM, Henry Saputra < > > > [email protected]> > > > > >> wrote: > > > > >> > > > > >>> Are taking VOTE on this or you just trying to do dry run? > > > > >>> > > > > >>> On Wed, Oct 8, 2014 at 4:15 AM, Robert Metzger < > > [email protected]> > > > > >>> wrote: > > > > >>> > Hi guys, > > > > >>> > > > > > >>> > I've created a first release candidate for the 0.7-incubating > > major > > > > >>> release: > > > > >>> > > > > > >>> > http://people.apache.org/~rmetzger/flink-0.7.0-incubating-rc0/ > > > > >>> > > > > > >>> > > > > > >>> > This thread is not an official release vote thread, rather a > > > > discussion > > > > >>> > thread to collect all open issues that we want to address until > > we > > > > >>> create > > > > >>> > the voting release candidate. > > > > >>> > > > > > >>> > > > > > >>> > Best, > > > > >>> > Robert > > > > >>> > > > > >> > > > > >> > > > > > > > > > >
