[DISCUSS] Dropping hadoop-accelerator download

[dsetrakyan]

Igniters,

I would like to suggest dropping the hadoop accelerator distribution of
Apache Ignite. Note, that it does not mean dropping any features. We are
still going to support IGFS, in-memory map reduce, spark integration, etc.
However, the hadoop-accelerator downloadable is putting extra pressure on
the community to test and document, which can be avoided. All the features
will still be supported through regular Ignite distribution.

Any objections?

D.

[Valentin Kulichenko]

+1 from me. Hadoop Accelerator build is confusing and doesn't provide any
value.

We already had a similar discussion couple of years back where I shared
some of my thoughts regarding this:
http://apache-ignite-developers.2346864.n4.nabble.com/ignite-spark-module-in-Hadoop-Accelerator-td12343.html

-Val

On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]>
wrote:

> Igniters,
>
> I would like to suggest dropping the hadoop accelerator distribution of
> Apache Ignite. Note, that it does not mean dropping any features. We are
> still going to support IGFS, in-memory map reduce, spark integration, etc.
> However, the hadoop-accelerator downloadable is putting extra pressure on
> the community to test and document, which can be avoided. All the features
> will still be supported through regular Ignite distribution.
>
> Any objections?
>
> D.
>

[Alexey Zinoviev]

+1
I asked a few guys, who use Ignite with Spark integration and  (sometimes)
IGFS, nobody use it via Hadoop Accelerator build

ср, 17 окт. 2018 г. в 4:12, Valentin Kulichenko <
[hidden email]>:

> +1 from me. Hadoop Accelerator build is confusing and doesn't provide any
> value.
>
> We already had a similar discussion couple of years back where I shared
> some of my thoughts regarding this:
>
> http://apache-ignite-developers.2346864.n4.nabble.com/ignite-spark-module-in-Hadoop-Accelerator-td12343.html
>
> -Val
>
> On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]>
> wrote:
>
> > Igniters,
> >
> > I would like to suggest dropping the hadoop accelerator distribution of
> > Apache Ignite. Note, that it does not mean dropping any features. We are
> > still going to support IGFS, in-memory map reduce, spark integration,
> etc.
> > However, the hadoop-accelerator downloadable is putting extra pressure on
> > the community to test and document, which can be avoided. All the
> features
> > will still be supported through regular Ignite distribution.
> >
> > Any objections?
> >
> > D.
> >
>

[dmagda]

No objections from my side.

--
Denis


On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]>
wrote:

> Igniters,
>
> I would like to suggest dropping the hadoop accelerator distribution of
> Apache Ignite. Note, that it does not mean dropping any features. We are
> still going to support IGFS, in-memory map reduce, spark integration, etc.
> However, the hadoop-accelerator downloadable is putting extra pressure on
> the community to test and document, which can be avoided. All the features
> will still be supported through regular Ignite distribution.
>
> Any objections?
>
> D.
>

[vveider]

Can we drop Hadoop binaries from delivery this release (2.7) or should it be moved to next?

> On 18 Oct 2018, at 04:38, Denis Magda <[hidden email]> wrote:
>
> No objections from my side.
>
> --
> Denis
>
>
> On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]>
> wrote:
>
>> Igniters,
>>
>> I would like to suggest dropping the hadoop accelerator distribution of
>> Apache Ignite. Note, that it does not mean dropping any features. We are
>> still going to support IGFS, in-memory map reduce, spark integration, etc.
>> However, the hadoop-accelerator downloadable is putting extra pressure on
>> the community to test and document, which can be avoided. All the features
>> will still be supported through regular Ignite distribution.
>>
>> Any objections?
>>
>> D.
>>

[dmagda]

I would do this if it doesn't impact the release time.

--
Denis

On Thu, Oct 18, 2018 at 2:34 AM Petr Ivanov <[hidden email]> wrote:

> Can we drop Hadoop binaries from delivery this release (2.7) or should it
> be moved to next?
>
>
> > On 18 Oct 2018, at 04:38, Denis Magda <[hidden email]> wrote:
> >
> > No objections from my side.
> >
> > --
> > Denis
> >
> >
> > On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]
> >
> > wrote:
> >
> >> Igniters,
> >>
> >> I would like to suggest dropping the hadoop accelerator distribution of
> >> Apache Ignite. Note, that it does not mean dropping any features. We are
> >> still going to support IGFS, in-memory map reduce, spark integration,
> etc.
> >> However, the hadoop-accelerator downloadable is putting extra pressure
> on
> >> the community to test and document, which can be avoided. All the
> features
> >> will still be supported through regular Ignite distribution.
> >>
> >> Any objections?
> >>
> >> D.
> >>
>
>

[vveider]

Can you file an issue then, please?

> On 19 Oct 2018, at 04:04, Denis Magda <[hidden email]> wrote:
>
> I would do this if it doesn't impact the release time.
>
> --
> Denis
>
> On Thu, Oct 18, 2018 at 2:34 AM Petr Ivanov <[hidden email]> wrote:
>
>> Can we drop Hadoop binaries from delivery this release (2.7) or should it
>> be moved to next?
>>
>>
>>> On 18 Oct 2018, at 04:38, Denis Magda <[hidden email]> wrote:
>>>
>>> No objections from my side.
>>>
>>> --
>>> Denis
>>>
>>>
>>> On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <[hidden email]
>>>
>>> wrote:
>>>
>>>> Igniters,
>>>>
>>>> I would like to suggest dropping the hadoop accelerator distribution of
>>>> Apache Ignite. Note, that it does not mean dropping any features. We are
>>>> still going to support IGFS, in-memory map reduce, spark integration,
>> etc.
>>>> However, the hadoop-accelerator downloadable is putting extra pressure
>> on
>>>> the community to test and document, which can be avoided. All the
>> features
>>>> will still be supported through regular Ignite distribution.
>>>>
>>>> Any objections?
>>>>
>>>> D.
>>>>
>>
>>

[dmagda]

Here it is: https://issues.apache.org/jira/browse/IGNITE-9953

--
Denis

On Fri, Oct 19, 2018 at 2:34 AM Petr Ivanov <[hidden email]> wrote:

> Can you file an issue then, please?
>
>
> > On 19 Oct 2018, at 04:04, Denis Magda <[hidden email]> wrote:
> >
> > I would do this if it doesn't impact the release time.
> >
> > --
> > Denis
> >
> > On Thu, Oct 18, 2018 at 2:34 AM Petr Ivanov <[hidden email]> wrote:
> >
> >> Can we drop Hadoop binaries from delivery this release (2.7) or should
> it
> >> be moved to next?
> >>
> >>
> >>> On 18 Oct 2018, at 04:38, Denis Magda <[hidden email]> wrote:
> >>>
> >>> No objections from my side.
> >>>
> >>> --
> >>> Denis
> >>>
> >>>
> >>> On Tue, Oct 16, 2018 at 4:45 PM Dmitriy Setrakyan <
> [hidden email]
> >>>
> >>> wrote:
> >>>
> >>>> Igniters,
> >>>>
> >>>> I would like to suggest dropping the hadoop accelerator distribution
> of
> >>>> Apache Ignite. Note, that it does not mean dropping any features. We
> are
> >>>> still going to support IGFS, in-memory map reduce, spark integration,
> >> etc.
> >>>> However, the hadoop-accelerator downloadable is putting extra pressure
> >> on
> >>>> the community to test and document, which can be avoided. All the
> >> features
> >>>> will still be supported through regular Ignite distribution.
> >>>>
> >>>> Any objections?
> >>>>
> >>>> D.
> >>>>
> >>
> >>
>
>