You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@reef.apache.org by Dongjoon Hyun <do...@apache.org> on 2015/05/08 04:37:18 UTC
Hi.
Hi, all.
I downloaded and tested rc6 before voting.
apache-reef-0.11.0-incubating-rc6.tar.gz
apache-reef-0.11.0-incubating-rc6.tar.gz.asc
apache-reef-0.11.0-incubating-rc6.tar.gz.md5
apache-reef-0.11.0-incubating-rc6.tar.gz.sha1
apache-reef-0.11.0-incubating-rc6.tar.gz.sha512
The tar file contains everything correctly with correct permission.
ASC, MD5, SHA1, SHA512 are correct, too.
Also, it passed 'mvn clean install' on my Mac.
But, it failed on CentOS 6.5, CDH 5.4, JDK 1.7.0_67 with the following
message.
( On the same cluster, latest REEF master branch is passed with mvn. )
---------
Failed tests:
FailDriverTest.testFailDriverAllocatedEvaluator:82->failOn:64
expected:<FAILED> but was:<FORCE_CLOSED>
Tests run: 39, Failures: 1, Errors: 0, Skipped: 0
[INFO]
------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO]
[INFO] REEF ............................................... SUCCESS [
0.983 s]
..... (skipped) ....
[INFO] Tang Test Jar B conflict A ......................... SUCCESS [
0.019 s]
[INFO] REEF Tests ......................................... FAILURE [07:50
min]
[INFO] REEF Website ....................................... SKIPPED
---------
This is just for a test result sharing, not something related with voting.
Warmly,
Dongjoon.
RE: Hi.
Posted by "Julia Wang (QIUHE)" <Qi...@microsoft.com>.
Thanks Dongjoon!
-----Original Message-----
From: Dongjoon Hyun [mailto:dongjoon@apache.org]
Sent: Friday, May 8, 2015 8:57 AM
To: dev@reef.incubator.apache.org
Subject: Re: Hi.
Hi, All.
According to the advice of Markus Weimer and Julia Wang, I downloaded RC6 newly and did the whole process again. And, now, it passed on both Mac and CentOS cluster a few minutes ago.
[hadoop@GPU-F02-01 ~]$ md5sum apache-reef-0.11.0-incubating-rc6.tar.gz
46a3fdd31df6d917a74650404ca28e2c apache-reef-0.11.0-incubating-rc6.tar.gz
It looks good.
Wamly,
Dongjoon Hyun.
On Sat, May 9, 2015 at 12:30 AM, Byung-Gon Chun <bg...@gmail.com> wrote:
> We can perhaps create a simple guideline: if we observe the same
> problem among two or more testers consistently, then we should fix the
> problem before releasing it.
>
> On Fri, May 8, 2015 at 7:40 AM, Markus Weimer <ma...@weimo.de> wrote:
>
> > Hi,
> >
> > > What is the policy with these sporadic test failures? I’m fine
> > > with releasing with them as known issues.
> >
> > We don't have one (yet), so let's make one up :-)
> >
> > I'm also in favor of releasing with known defects. The code is
> > vastly improved since 0.10, which probably also contained the very same issues.
> >
> > However, I have hard time turning this on-the-spot view into a "policy".
> > Hence I propose not to have one. Instead, let's play it by ear with
> > each release.
> >
> > Sounds good?
> >
> > Markus
> >
>
>
>
> --
> Byung-Gon Chun
>
Re: Hi.
Posted by Dongjoon Hyun <do...@apache.org>.
Hi, All.
According to the advice of Markus Weimer and Julia Wang, I downloaded RC6
newly and did the whole process again. And, now, it passed on both Mac and
CentOS cluster a few minutes ago.
[hadoop@GPU-F02-01 ~]$ md5sum apache-reef-0.11.0-incubating-rc6.tar.gz
46a3fdd31df6d917a74650404ca28e2c apache-reef-0.11.0-incubating-rc6.tar.gz
It looks good.
Wamly,
Dongjoon Hyun.
On Sat, May 9, 2015 at 12:30 AM, Byung-Gon Chun <bg...@gmail.com> wrote:
> We can perhaps create a simple guideline: if we observe the same problem
> among two or more testers consistently, then we should fix the problem
> before releasing it.
>
> On Fri, May 8, 2015 at 7:40 AM, Markus Weimer <ma...@weimo.de> wrote:
>
> > Hi,
> >
> > > What is the policy with these sporadic test failures? I’m fine with
> > > releasing with them as known issues.
> >
> > We don't have one (yet), so let's make one up :-)
> >
> > I'm also in favor of releasing with known defects. The code is vastly
> > improved since 0.10, which probably also contained the very same issues.
> >
> > However, I have hard time turning this on-the-spot view into a "policy".
> > Hence I propose not to have one. Instead, let's play it by ear with each
> > release.
> >
> > Sounds good?
> >
> > Markus
> >
>
>
>
> --
> Byung-Gon Chun
>
Re: Hi.
Posted by Byung-Gon Chun <bg...@gmail.com>.
We can perhaps create a simple guideline: if we observe the same problem
among two or more testers consistently, then we should fix the problem
before releasing it.
On Fri, May 8, 2015 at 7:40 AM, Markus Weimer <ma...@weimo.de> wrote:
> Hi,
>
> > What is the policy with these sporadic test failures? I’m fine with
> > releasing with them as known issues.
>
> We don't have one (yet), so let's make one up :-)
>
> I'm also in favor of releasing with known defects. The code is vastly
> improved since 0.10, which probably also contained the very same issues.
>
> However, I have hard time turning this on-the-spot view into a "policy".
> Hence I propose not to have one. Instead, let's play it by ear with each
> release.
>
> Sounds good?
>
> Markus
>
--
Byung-Gon Chun
Re: Hi.
Posted by Markus Weimer <ma...@weimo.de>.
Hi,
> What is the policy with these sporadic test failures? I’m fine with
> releasing with them as known issues.
We don't have one (yet), so let's make one up :-)
I'm also in favor of releasing with known defects. The code is vastly
improved since 0.10, which probably also contained the very same issues.
However, I have hard time turning this on-the-spot view into a "policy".
Hence I propose not to have one. Instead, let's play it by ear with each
release.
Sounds good?
Markus
RE: Hi.
Posted by Brian Cho <ch...@gmail.com>.
Thanks for the testing Dongjoon :)
My guess is that this is the same issue I ran into here: https://issues.apache.org/jira/browse/REEF-294
What is the policy with these sporadic test failures? I’m fine with releasing with them as known issues.
-Brian
On May 8, 2015 at 4:02:20 PM, Julia Wang (QIUHE) (qiuhe.wang@microsoft.com) wrote:
Thanks Dongjoon for helping on the testing!
I did fresh download this afternoon from apache website after I uploaded RC6 from my another machine, I untar, build and run tests for both java and .Net on Windows. They all passed.
With the same source code, Maven tests failed once when I did dry run last week. But in all the other runs, they all passed. May be that was caused by one of the one of the flukes.
If you can run again, or someone else has similar env and try again, that would help.
Thanks,
Julia
-----Original Message-----
From: Markus Weimer [mailto:markus@weimo.de]
Sent: Thursday, May 07, 2015 9:37 PM
To: dev@reef.incubator.apache.org
Subject: Re: Hi.
Hi,
thanks for testing! Can you run the same tests again to see whether it is one of the flukes we have been chasing?
Markus
On 2015-05-07 19:37, Dongjoon Hyun wrote:
> Hi, all.
>
> I downloaded and tested rc6 before voting.
>
> apache-reef-0.11.0-incubating-rc6.tar.gz
> apache-reef-0.11.0-incubating-rc6.tar.gz.asc
> apache-reef-0.11.0-incubating-rc6.tar.gz.md5
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha1
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha512
>
> The tar file contains everything correctly with correct permission.
> ASC, MD5, SHA1, SHA512 are correct, too.
> Also, it passed 'mvn clean install' on my Mac.
>
> But, it failed on CentOS 6.5, CDH 5.4, JDK 1.7.0_67 with the following
> message.
> ( On the same cluster, latest REEF master branch is passed with mvn. )
>
> ---------
> Failed tests:
> FailDriverTest.testFailDriverAllocatedEvaluator:82->failOn:64
> expected:<FAILED> but was:<FORCE_CLOSED>
>
> Tests run: 39, Failures: 1, Errors: 0, Skipped: 0
>
> [INFO]
> ----------------------------------------------------------------------
> --
> [INFO] Reactor Summary:
> [INFO]
> [INFO] REEF ............................................... SUCCESS [
> 0.983 s]
> ..... (skipped) ....
> [INFO] Tang Test Jar B conflict A ......................... SUCCESS [
> 0.019 s]
> [INFO] REEF Tests ......................................... FAILURE
> [07:50 min] [INFO] REEF Website
> ....................................... SKIPPED
> ---------
>
> This is just for a test result sharing, not something related with voting.
>
> Warmly,
> Dongjoon.
>
RE: Hi.
Posted by "Julia Wang (QIUHE)" <Qi...@microsoft.com>.
Thanks Dongjoon for helping on the testing!
I did fresh download this afternoon from apache website after I uploaded RC6 from my another machine, I untar, build and run tests for both java and .Net on Windows. They all passed.
With the same source code, Maven tests failed once when I did dry run last week. But in all the other runs, they all passed. May be that was caused by one of the one of the flukes.
If you can run again, or someone else has similar env and try again, that would help.
Thanks,
Julia
-----Original Message-----
From: Markus Weimer [mailto:markus@weimo.de]
Sent: Thursday, May 07, 2015 9:37 PM
To: dev@reef.incubator.apache.org
Subject: Re: Hi.
Hi,
thanks for testing! Can you run the same tests again to see whether it is one of the flukes we have been chasing?
Markus
On 2015-05-07 19:37, Dongjoon Hyun wrote:
> Hi, all.
>
> I downloaded and tested rc6 before voting.
>
> apache-reef-0.11.0-incubating-rc6.tar.gz
> apache-reef-0.11.0-incubating-rc6.tar.gz.asc
> apache-reef-0.11.0-incubating-rc6.tar.gz.md5
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha1
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha512
>
> The tar file contains everything correctly with correct permission.
> ASC, MD5, SHA1, SHA512 are correct, too.
> Also, it passed 'mvn clean install' on my Mac.
>
> But, it failed on CentOS 6.5, CDH 5.4, JDK 1.7.0_67 with the following
> message.
> ( On the same cluster, latest REEF master branch is passed with mvn. )
>
> ---------
> Failed tests:
> FailDriverTest.testFailDriverAllocatedEvaluator:82->failOn:64
> expected:<FAILED> but was:<FORCE_CLOSED>
>
> Tests run: 39, Failures: 1, Errors: 0, Skipped: 0
>
> [INFO]
> ----------------------------------------------------------------------
> --
> [INFO] Reactor Summary:
> [INFO]
> [INFO] REEF ............................................... SUCCESS [
> 0.983 s]
> ..... (skipped) ....
> [INFO] Tang Test Jar B conflict A ......................... SUCCESS [
> 0.019 s]
> [INFO] REEF Tests ......................................... FAILURE
> [07:50 min] [INFO] REEF Website
> ....................................... SKIPPED
> ---------
>
> This is just for a test result sharing, not something related with voting.
>
> Warmly,
> Dongjoon.
>
Re: Hi.
Posted by Markus Weimer <ma...@weimo.de>.
Hi,
thanks for testing! Can you run the same tests again to see whether it
is one of the flukes we have been chasing?
Markus
On 2015-05-07 19:37, Dongjoon Hyun wrote:
> Hi, all.
>
> I downloaded and tested rc6 before voting.
>
> apache-reef-0.11.0-incubating-rc6.tar.gz
> apache-reef-0.11.0-incubating-rc6.tar.gz.asc
> apache-reef-0.11.0-incubating-rc6.tar.gz.md5
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha1
> apache-reef-0.11.0-incubating-rc6.tar.gz.sha512
>
> The tar file contains everything correctly with correct permission.
> ASC, MD5, SHA1, SHA512 are correct, too.
> Also, it passed 'mvn clean install' on my Mac.
>
> But, it failed on CentOS 6.5, CDH 5.4, JDK 1.7.0_67 with the following
> message.
> ( On the same cluster, latest REEF master branch is passed with mvn. )
>
> ---------
> Failed tests:
> FailDriverTest.testFailDriverAllocatedEvaluator:82->failOn:64
> expected:<FAILED> but was:<FORCE_CLOSED>
>
> Tests run: 39, Failures: 1, Errors: 0, Skipped: 0
>
> [INFO]
> ------------------------------------------------------------------------
> [INFO] Reactor Summary:
> [INFO]
> [INFO] REEF ............................................... SUCCESS [
> 0.983 s]
> ..... (skipped) ....
> [INFO] Tang Test Jar B conflict A ......................... SUCCESS [
> 0.019 s]
> [INFO] REEF Tests ......................................... FAILURE [07:50
> min]
> [INFO] REEF Website ....................................... SKIPPED
> ---------
>
> This is just for a test result sharing, not something related with voting.
>
> Warmly,
> Dongjoon.
>