Andrew, You are right, there are paths that could make that test pass, fail or even hang, based on timing. I had written the test without completely understanding the RPC thread pool usage. Please open a JIRA for that test, and add a @Ignore annotation with a TODO(DRILL-<jira_number>). I'll add comments to that JIRA.
- Sudheesh > On Oct 21, 2015, at 5:34 PM, andrew <and...@primer.org> wrote: > > I documented the failure on the JIRA. I don’t think there is more information > as no exception was thrown. > >> On Oct 21, 2015, at 5:30 PM, Abdel Hakim Deneche <adene...@maprtech.com> >> wrote: >> >> what kind of failures are you seeing ? >> >> On Wed, Oct 21, 2015 at 4:09 PM, andrew <and...@primer.org> wrote: >> >>> hey drillers - >>> >>> i’ve been blocked trying to get our hadoop dependency upgraded to 2.7.1 ( >>> https://issues.apache.org/jira/browse/DRILL-3749 < >>> https://issues.apache.org/jira/browse/DRILL-3749>). having trouble with >>> TestDrillbitResilience.cancelAfterEverythingIsCompleted(). >>> >>> the test runs fine on my laptop, but fails on an AWS m2.xlarge instance. >>> given that, i’m wondering if it’s a timing issue that just doesn’t manifest >>> on underpowered machines. >>> >>> is there anyone available to take a look? >>> >>> thx. >>> andrew >>> >>> >>> >> >> >> -- >> >> Abdelhakim Deneche >> >> Software Engineer >> >> <http://www.mapr.com/> >> >> >> Now Available - Free Hadoop On-Demand Training >> <http://www.mapr.com/training?utm_source=Email&utm_medium=Signature&utm_campaign=Free%20available> >
