ignite-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dmitriy Pavlov <dpavlov....@gmail.com>
Subject Re: [Discussion] revert of commit MVCC, ignite-9320
Date Fri, 28 Sep 2018 07:35:08 GMT
Hi Dmitriy,

Why not revert the change?

This test failure was appropriately reported to the dev list, and the
contributor did not fix it:
http://apache-ignite-developers.2346864.n4.nabble.com/MTCGA-new-failures-in-builds-1888723-needs-to-be-handled-td35239.html


It seems this idea of filing blockers does not work well.  It means for
community member:
- someday/maybe/someone will fix this blocker
- or he/she will move it to 3.0
- or we will reclassify and minor and mute test.

I can see tons of tests fixes going from 2.7 to 2.8. It always means I, for
example, will break test today, and someone will come and do not-so-fun
work for me.

I see tons of tickets about fixing the tests, so I prefer that Igniters who
invest their time here should achieve the result (green master), but not
those who are ignoring How To Contribute, even it is GridGain'ers.

I suggest reverts of new stable test failures should be more or less
automatic after 72h as a reasonable time to fix it. If contributor would
like to finalize solution (and most of the cases it is so), a test will be
fixed in 3 days.

If not, why to have not ready solutions in master, let's have it in
separate branches.

So, what do you suggest instead?

Sincerely,
Dmitriy Pavlov

пт, 28 сент. 2018 г. в 2:40, Dmitriy Setrakyan <dsetrakyan@apache.org>:

> Let's not revert any commits yet. Can we find out who did the commit and
> why he/she is not fixing the test?
>
> D.
>
> On Thu, Sep 27, 2018 at 4:21 PM Vyacheslav Daradur <daradurvs@gmail.com>
> wrote:
>
> > Hi,
> >
> > Are you talking about
> > 'IgniteConfigurationParityTest#TestIgniteConfiguration'?
> >
> > Seems it's not hard to fix this test, it's necessary just to implement
> > missing members (at least as stubs) on .NET side in
> > IgniteConfiguration class.
> >
> > Is there a Jira issue?
> >
> > On Fri, Sep 28, 2018 at 2:12 AM Dmitriy Pavlov <dpavlov.spb@gmail.com>
> > wrote:
> > >
> > > Hi,
> > >
> > > I'm grateful for contributions made in that area, but it seems folks
> > don't
> > > have time to fix the test.
> > >
> > >
> > > Tomorrow I'm going to revert commit.
> > >
> > > It seems it is the only way we can keep master more or less green.
> > >
> > >
> >
> https://ci.ignite.apache.org/viewLog.html?buildId=1888723&tab=buildChangesDiv&buildTypeId=IgniteTests24Java8_PlatformNet
> > >
> > >
> > > Sincerely
> > > Dmitry Pavlov
> >
> >
> >
> > --
> > Best Regards, Vyacheslav D.
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message