with some more latches cancel() can be 'sync' to be deterministic and not depending on time I think, isn't it?
Romain Manni-Bucau Twitter: @rmannibucau Blog: http://rmannibucau.wordpress.com/ LinkedIn: http://fr.linkedin.com/in/rmannibucau Github: https://github.com/rmannibucau 2014-06-04 15:44 GMT+02:00 Andy Gumbrecht <agumbre...@tomitribe.com>: > Sure, we agree. > > I am looking at the tests now and will commit something better soon. > The test is also testing the ability to cancel the scheduled task, so it is > not really important that the 'count' is accurate. > It does however have to test over an arbitrary period of time. > > Andy. > > On 04/06/2014 14:31, hanas...@gmail.com wrote: >> >> *From:* Romain Manni-Bucau <rmannibu...@gmail.com> >> *Sent:* Wednesday, June 04, 2014 6:48AM >> *To:* dev@tomee.apache.org <dev@tomee.apache.org> >> *Subject:* Re: Fwd: svn commit: r1600089 - >> /tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >> >> Is this based on my posting yesterday about a build failure that >> subsequently passed when increasing the sleep time? >> >> Either way, the thread waits are a bit indeterminate and could lead to >> longer build times and unrepeatable success / failure builds based on the >> size of hardware or load at the time of build. >> >> On 06/04/2014 06:48 AM, Romain Manni-Bucau wrote: >>> >>> Hehe no hurry ;). >>> >>> >>> Romain Manni-Bucau >>> Twitter: @rmannibucau >>> Blog: http://rmannibucau.wordpress.com/ >>> LinkedIn: http://fr.linkedin.com/in/rmannibucau >>> Github: https://github.com/rmannibucau >>> >>> >>> 2014-06-04 13:44 GMT+02:00 Andy Gumbrecht <agumbre...@tomitribe.com>: >>>> >>>> Sure, sounds good to me - I'll have another look in a minute, just got >>>> to >>>> get a sandwich for lunch :-D >>>> >>>> Andy. >>>> >>>> >>>> On 04/06/2014 13:39, Romain Manni-Bucau wrote: >>>>> >>>>> >>>>> I'd go for a version with latches and no sleep rather. >>>>> >>>>> >>>>> Romain Manni-Bucau >>>>> Twitter: @rmannibucau >>>>> Blog: http://rmannibucau.wordpress.com/ >>>>> LinkedIn: http://fr.linkedin.com/in/rmannibucau >>>>> Github: https://github.com/rmannibucau >>>>> >>>>> >>>>> 2014-06-04 13:28 GMT+02:00 Andy Gumbrecht <agumbre...@tomitribe.com>: >>>>>> >>>>>> >>>>>> I saw that also, and decided to leave it - I.e. not change it. >>>>>> >>>>>> It looks like it's there because the test could potentially return 7 >>>>>> due >>>>>> to >>>>>> getNextRunTime and Thread.sleep(5000) not really guaranteeing count >>>>>> accuracy >>>>>> . >>>>>> >>>>>> Maybe just testing for > 5 is better. >>>>>> >>>>>> Andy. >>>>>> >>>>>> >>>>>> On 04/06/2014 13:06, Romain Manni-Bucau wrote: >>>>>>> >>>>>>> >>>>>>> any reason to tolerate 1 of delta? >>>>>>> >>>>>>> >>>>>>> Romain Manni-Bucau >>>>>>> Twitter: @rmannibucau >>>>>>> Blog: http://rmannibucau.wordpress.com/ >>>>>>> LinkedIn: http://fr.linkedin.com/in/rmannibucau >>>>>>> Github: https://github.com/rmannibucau >>>>>>> >>>>>>> >>>>>>> >>>>>>> ---------- Forwarded message ---------- >>>>>>> From: <andygumbre...@apache.org> >>>>>>> Date: 2014-06-04 12:26 GMT+02:00 >>>>>>> Subject: svn commit: r1600089 - >>>>>>> >>>>>>> >>>>>>> >>>>>>> /tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >>>>>>> To: comm...@tomee.apache.org >>>>>>> >>>>>>> >>>>>>> Author: andygumbrecht >>>>>>> Date: Wed Jun 4 10:26:12 2014 >>>>>>> New Revision: 1600089 >>>>>>> >>>>>>> URL: http://svn.apache.org/r1600089 >>>>>>> Log: >>>>>>> Use a blocking method >>>>>>> >>>>>>> Modified: >>>>>>> >>>>>>> >>>>>>> >>>>>>> tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >>>>>>> >>>>>>> Modified: >>>>>>> >>>>>>> >>>>>>> tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >>>>>>> URL: >>>>>>> >>>>>>> >>>>>>> http://svn.apache.org/viewvc/tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java?rev=1600089&r1=1600088&r2=1600089&view=diff >>>>>>> >>>>>>> >>>>>>> >>>>>>> ============================================================================== >>>>>>> --- >>>>>>> >>>>>>> >>>>>>> tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >>>>>>> (original) >>>>>>> +++ >>>>>>> >>>>>>> >>>>>>> tomee/tomee/trunk/container/openejb-concurrency-utilities-ee/src/test/java/org/apache/openejb/concurrencyutilities/test/ManagedScheduledExecutorServiceTest.java >>>>>>> Wed Jun 4 10:26:12 2014 >>>>>>> @@ -81,7 +81,7 @@ public class ManagedScheduledExecutorSer >>>>>>> assertEquals(6, future.get().intValue()); >>>>>>> >>>>>>> future.cancel(true); >>>>>>> - assertEquals(6, counter.get(), 1); >>>>>>> + assertEquals(6, counter.getAndIncrement(), 1); >>>>>>> >>>>>>> Thread.sleep(2000); // since get() is not blocking, wait >>>>>>> a >>>>>>> bit the task ends up >>>>>>> >>>>>>> @@ -118,7 +118,7 @@ public class ManagedScheduledExecutorSer >>>>>>> Thread.sleep(5000); >>>>>>> >>>>>>> future.cancel(true); >>>>>>> - assertEquals(6, counter.get(), 1); >>>>>>> + assertEquals(6, counter.getAndIncrement(), 1); >>>>>>> >>>>>>> Thread.sleep(2000); // since get() is not blocking, wait >>>>>>> a >>>>>>> bit the task ends >>>>>>> >>>>>>> >>>>>> -- >>>>>> Andy Gumbrecht >>>>>> >>>>>> http://www.tomitribe.com >>>>>> agumbre...@tomitribe.com >>>>>> https://twitter.com/AndyGeeDe >>>>>> >>>>>> TomEE treibt Tomitribe! | http://tomee.apache.org >>>>>> >>>>> >>>> >>>> -- >>>> Andy Gumbrecht >>>> >>>> http://www.tomitribe.com >>>> agumbre...@tomitribe.com >>>> https://twitter.com/AndyGeeDe >>>> >>>> TomEE treibt Tomitribe! | http://tomee.apache.org >>>> >> >> > > -- > Andy Gumbrecht > > http://www.tomitribe.com > agumbre...@tomitribe.com > https://twitter.com/AndyGeeDe > > TomEE treibt Tomitribe! | http://tomee.apache.org >