Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
HBase, mail # dev - State of the 0.94 tests


+
lars hofhansl 2012-10-07, 06:11
+
lars hofhansl 2012-10-07, 06:41
+
Andrew Purtell 2012-10-07, 08:36
+
lars hofhansl 2012-10-07, 16:21
+
lars hofhansl 2012-10-07, 16:26
+
lars hofhansl 2012-10-07, 16:44
+
lars hofhansl 2012-10-07, 18:23
+
lars hofhansl 2012-10-07, 20:36
+
Ramkrishna.S.Vasudevan 2012-10-08, 04:01
Copy link to this message
-
Re: State of the 0.94 tests
lars hofhansl 2012-10-08, 04:15
Thanks Ram,

as I said, these are well thought out fixes. Sometimes I think you are the only one who actually understands how the assignment/balance process really works :)
With the environment issues out of the way (by not using ubuntu2) the test fails much less frequently.
The frequently failures were actually my fault by enabling ubuntu2 for the jenkins build in the first place (there was a reason why someone had disabled it before).
testShouldThrowIOExceptionIfStoreFileSizeIsEmptyAndSHouldSuccessfullyExecuteRollback still fails occasionally with a failed assertion, but these look like test failures, not production code problems. I commented to this extend on the jira.

TL;DR: I think we're good for the current RC. The test flaps sometimes; we just need to fix that.

-- Lars

________________________________
 From: Ramkrishna.S.Vasudevan <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]; 'lars hofhansl' <[EMAIL PROTECTED]>
Sent: Sunday, October 7, 2012 9:01 PM
Subject: RE: State of the 0.94 tests
 
Hi Lars

I was not in town and was in travel for the last 2 days.  I will immediately
check the reason for the testcase failures.  Had I been there I would have
helped out earlier.

Sorry about that.

Regards
Ram

> -----Original Message-----
> From: lars hofhansl [mailto:[EMAIL PROTECTED]]
> Sent: Monday, October 08, 2012 2:07 AM
> To: [EMAIL PROTECTED]
> Subject: Re: State of the 0.94 tests
>
> After this change things look better. Apologies for the noise. Stay
> tuned for the next RC.
>
> -- Lars
>
>
>
> ________________________________
>  From: lars hofhansl <[EMAIL PROTECTED]>
> To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
> Sent: Sunday, October 7, 2012 11:23 AM
> Subject: Re: State of the 0.94 tests
>
> I looked back through the failures. I had recently enabled all "ubuntu"
> build vms for the 0.94 builds.
> It turns out that most of the environment issues occur on ubuntu2. I
> excluded that from the build vms.
>
>
> -- Lars
>
>
>
> ________________________________
> From: Andrew Purtell <[EMAIL PROTECTED]>
> To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
> Sent: Sunday, October 7, 2012 1:36 AM
> Subject: Re: State of the 0.94 tests
>
> Too many open files usually is an environment issue.
>
> Lars, you should consider setting up a private Jenkins as a sanity
> check.
>
> On Oct 7, 2012, at 2:41 PM, lars hofhansl <[EMAIL PROTECTED]> wrote:
>
> > Looks like after all that whining I finally got a successful build.
> > But I lost confidence in the current 0.94 code line.
> >
> > Still, it is possible that all of these were environmental issue. If
> we can get a few more successful runs, it could be OK.
> >
> > -- Lars
> >
> >
> >
> > ________________________________
> > From: lars hofhansl <[EMAIL PROTECTED]>
> > To: hbase-dev <[EMAIL PROTECTED]>
> > Sent: Saturday, October 6, 2012 11:11 PM
> > Subject: State of the 0.94 tests
> >
> > I've been trying (essentially the entire day) getting a successful
> jenkins build for 0.94 (triggering the test run periodically from my
> phone). Not a *single* run succeeded.
> > This is clearly not acceptable. Something is off.
> >
> > The tests that fails the most frequently are:
> > -
> TestSplitTransactionOnCluster.testShouldThrowIOExceptionIfStoreFileSize
> IsEmptyAndSHouldSuccessfullyExecuteRollback
> > -
> TestSplitTransactionOnCluster.testShouldClearRITWhenNodeFoundInSplittin
> gState
> > (The failure cause most of the time is too many files open, but also
> fail because of unavailable regions).
> >
> > Both tests were added recently (since 0.94.2RC2). See HBASE-6854 and
> HBASE-6853.
> >
> > Either there is something wrong with the tests, or we introduced some
> problems in the code base.
> >
> > Note that I am not dinging these two changes specifically. Both were
> fixes with a lot of thought and care behind them.
> >
> > There are also various time out issues in other tests.
> >
> > These were all the fixes added since the last RC:
+
Andrew Purtell 2012-10-07, 08:46
+
Andrew Purtell 2012-10-07, 08:48