You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mrunit.apache.org by Konstantin Boudnik <co...@apache.org> on 2016/03/15 20:33:49 UTC
[DISCUSS] Moving MRUnit to attic
Hello.
I want to start the discussion about the near future of the project. Here's my
line of thinking. The MapReduce paradigm is clearly dying out and is getting
replaced more and more either by new MapReduce-like platforms (think Spark) or
by direct streaming processing and in-memory systems, completely going around
the batch.
MRUnit is definitely a good project to help developers to deal with MR
realities before the code hits the cluster. But evidently there isn't much
activity going on as dev@ list traffic (literally <10 JIRA messages/month) [1]
and line of releases indicate. According to reporter.apache.org the last
release had happened on May 2012, but I found some email evidence that 1.1.0
was out in June 2014 [2](?)
Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame in
it. If anything, it is a part of the natural cycle of any software project.
And most important, the code base will not go anywhere and can be resurrected
any time if there's a community to work on it.
Thoughts?
Cos
1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
2. http://is.gd/mgFXsC
3. http://attic.apache.org/
Re: [DISCUSS] Moving MRUnit to attic
Posted by Dave Beech <db...@apache.org>.
>From the webpage: "*Projects whose PMC are unable to muster 3 votes for a
release, who have no active committers or are unable to fulfill their
reporting duties to the board are all good candidates for the Attic*."
I think, unfortunately, MRUnit is hitting all three of those points. So it
would have to be +1 from me.
Dave
On 16 March 2016 at 01:33, Patrick Hunt <ph...@apache.org> wrote:
> Background/implications here: https://attic.apache.org/
>
> Patrick
>
> On Tue, Mar 15, 2016 at 6:28 PM, Ryan Chapin <rc...@nbinteractive.com>
> wrote:
> > There are still a lot of companies that have active, MapReduce codebases
> > that are using MRUnit. Not to say that newer technologies such as Spark
> > aren't becoming more popular but there are still instances when the tried
> > and true MapReduce paradigm is the choice for certain projects and makes
> the
> > most sense.
> >
> > Honestly, I am new to this project and not exactly sure what moving
> > something to the attic means or what the consequences to the project will
> > be. I got involved in it by I submitting a patch back in November for a
> fix
> > that I have wanted for quite some time. I would actually like to see it
> > integrated into a release as I use that MR feature on a regular basis.
> >
> > I would like to see it stay available and receive bug fixes when
> necessary.
> >
> > -- Ryan Chapin
> >
> >
> > On 03/15/2016 04:24 PM, Patrick Hunt wrote:
> >>
> >> Just because it's in the attic doesn't mean it's gone, just not under
> >> active development, which seems fine to me. Still available for folks
> >> to use.
> >>
> >> Patrick
> >>
> >> On Tue, Mar 15, 2016 at 12:41 PM, Cosmin Lehene <cl...@adobe.com>
> wrote:
> >>>
> >>> MRUnit has been a useful tool.
> >>> I'm glad to have benefited from and contribute to it, so I'd like to
> >>> thank the community!
> >>>
> >>> We're not writing new stuff in MR and instead use alternatives (like
> >>> Spark) so I think this would make sense.
> >>>
> >>> +1
> >>>
> >>> Cosmin
> >>>
> >>> ________________________________________
> >>> From: Konstantin Boudnik <co...@apache.org>
> >>> Sent: Tuesday, March 15, 2016 12:33 PM
> >>> To: dev@mrunit.apache.org
> >>> Subject: [DISCUSS] Moving MRUnit to attic
> >>>
> >>> Hello.
> >>>
> >>> I want to start the discussion about the near future of the project.
> >>> Here's my
> >>> line of thinking. The MapReduce paradigm is clearly dying out and is
> >>> getting
> >>> replaced more and more either by new MapReduce-like platforms (think
> >>> Spark) or
> >>> by direct streaming processing and in-memory systems, completely going
> >>> around
> >>> the batch.
> >>>
> >>> MRUnit is definitely a good project to help developers to deal with MR
> >>> realities before the code hits the cluster. But evidently there isn't
> >>> much
> >>> activity going on as dev@ list traffic (literally <10 JIRA
> >>> messages/month) [1]
> >>> and line of releases indicate. According to reporter.apache.org the
> last
> >>> release had happened on May 2012, but I found some email evidence that
> >>> 1.1.0
> >>> was out in June 2014 [2](?)
> >>>
> >>> Hence, is it a time perhaps to move MRUnit to attic [3]? There's no
> shame
> >>> in
> >>> it. If anything, it is a part of the natural cycle of any software
> >>> project.
> >>> And most important, the code base will not go anywhere and can be
> >>> resurrected
> >>> any time if there's a community to work on it.
> >>>
> >>> Thoughts?
> >>> Cos
> >>>
> >>> 1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
> >>> 2. http://is.gd/mgFXsC
> >>> 3. http://attic.apache.org/
> >>>
> >
> > --
> > Ryan Chapin
> >
>
Re: [DISCUSS] Moving MRUnit to attic
Posted by Patrick Hunt <ph...@apache.org>.
Background/implications here: https://attic.apache.org/
Patrick
On Tue, Mar 15, 2016 at 6:28 PM, Ryan Chapin <rc...@nbinteractive.com> wrote:
> There are still a lot of companies that have active, MapReduce codebases
> that are using MRUnit. Not to say that newer technologies such as Spark
> aren't becoming more popular but there are still instances when the tried
> and true MapReduce paradigm is the choice for certain projects and makes the
> most sense.
>
> Honestly, I am new to this project and not exactly sure what moving
> something to the attic means or what the consequences to the project will
> be. I got involved in it by I submitting a patch back in November for a fix
> that I have wanted for quite some time. I would actually like to see it
> integrated into a release as I use that MR feature on a regular basis.
>
> I would like to see it stay available and receive bug fixes when necessary.
>
> -- Ryan Chapin
>
>
> On 03/15/2016 04:24 PM, Patrick Hunt wrote:
>>
>> Just because it's in the attic doesn't mean it's gone, just not under
>> active development, which seems fine to me. Still available for folks
>> to use.
>>
>> Patrick
>>
>> On Tue, Mar 15, 2016 at 12:41 PM, Cosmin Lehene <cl...@adobe.com> wrote:
>>>
>>> MRUnit has been a useful tool.
>>> I'm glad to have benefited from and contribute to it, so I'd like to
>>> thank the community!
>>>
>>> We're not writing new stuff in MR and instead use alternatives (like
>>> Spark) so I think this would make sense.
>>>
>>> +1
>>>
>>> Cosmin
>>>
>>> ________________________________________
>>> From: Konstantin Boudnik <co...@apache.org>
>>> Sent: Tuesday, March 15, 2016 12:33 PM
>>> To: dev@mrunit.apache.org
>>> Subject: [DISCUSS] Moving MRUnit to attic
>>>
>>> Hello.
>>>
>>> I want to start the discussion about the near future of the project.
>>> Here's my
>>> line of thinking. The MapReduce paradigm is clearly dying out and is
>>> getting
>>> replaced more and more either by new MapReduce-like platforms (think
>>> Spark) or
>>> by direct streaming processing and in-memory systems, completely going
>>> around
>>> the batch.
>>>
>>> MRUnit is definitely a good project to help developers to deal with MR
>>> realities before the code hits the cluster. But evidently there isn't
>>> much
>>> activity going on as dev@ list traffic (literally <10 JIRA
>>> messages/month) [1]
>>> and line of releases indicate. According to reporter.apache.org the last
>>> release had happened on May 2012, but I found some email evidence that
>>> 1.1.0
>>> was out in June 2014 [2](?)
>>>
>>> Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame
>>> in
>>> it. If anything, it is a part of the natural cycle of any software
>>> project.
>>> And most important, the code base will not go anywhere and can be
>>> resurrected
>>> any time if there's a community to work on it.
>>>
>>> Thoughts?
>>> Cos
>>>
>>> 1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
>>> 2. http://is.gd/mgFXsC
>>> 3. http://attic.apache.org/
>>>
>
> --
> Ryan Chapin
>
Re: [DISCUSS] Moving MRUnit to attic
Posted by Ryan Chapin <rc...@nbinteractive.com>.
There are still a lot of companies that have active, MapReduce codebases
that are using MRUnit. Not to say that newer technologies such as Spark
aren't becoming more popular but there are still instances when the
tried and true MapReduce paradigm is the choice for certain projects and
makes the most sense.
Honestly, I am new to this project and not exactly sure what moving
something to the attic means or what the consequences to the project
will be. I got involved in it by I submitting a patch back in November
for a fix that I have wanted for quite some time. I would actually like
to see it integrated into a release as I use that MR feature on a
regular basis.
I would like to see it stay available and receive bug fixes when necessary.
-- Ryan Chapin
On 03/15/2016 04:24 PM, Patrick Hunt wrote:
> Just because it's in the attic doesn't mean it's gone, just not under
> active development, which seems fine to me. Still available for folks
> to use.
>
> Patrick
>
> On Tue, Mar 15, 2016 at 12:41 PM, Cosmin Lehene <cl...@adobe.com> wrote:
>> MRUnit has been a useful tool.
>> I'm glad to have benefited from and contribute to it, so I'd like to thank the community!
>>
>> We're not writing new stuff in MR and instead use alternatives (like Spark) so I think this would make sense.
>>
>> +1
>>
>> Cosmin
>>
>> ________________________________________
>> From: Konstantin Boudnik <co...@apache.org>
>> Sent: Tuesday, March 15, 2016 12:33 PM
>> To: dev@mrunit.apache.org
>> Subject: [DISCUSS] Moving MRUnit to attic
>>
>> Hello.
>>
>> I want to start the discussion about the near future of the project. Here's my
>> line of thinking. The MapReduce paradigm is clearly dying out and is getting
>> replaced more and more either by new MapReduce-like platforms (think Spark) or
>> by direct streaming processing and in-memory systems, completely going around
>> the batch.
>>
>> MRUnit is definitely a good project to help developers to deal with MR
>> realities before the code hits the cluster. But evidently there isn't much
>> activity going on as dev@ list traffic (literally <10 JIRA messages/month) [1]
>> and line of releases indicate. According to reporter.apache.org the last
>> release had happened on May 2012, but I found some email evidence that 1.1.0
>> was out in June 2014 [2](?)
>>
>> Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame in
>> it. If anything, it is a part of the natural cycle of any software project.
>> And most important, the code base will not go anywhere and can be resurrected
>> any time if there's a community to work on it.
>>
>> Thoughts?
>> Cos
>>
>> 1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
>> 2. http://is.gd/mgFXsC
>> 3. http://attic.apache.org/
>>
--
Ryan Chapin
Re: [DISCUSS] Moving MRUnit to attic
Posted by Patrick Hunt <ph...@apache.org>.
Just because it's in the attic doesn't mean it's gone, just not under
active development, which seems fine to me. Still available for folks
to use.
Patrick
On Tue, Mar 15, 2016 at 12:41 PM, Cosmin Lehene <cl...@adobe.com> wrote:
> MRUnit has been a useful tool.
> I'm glad to have benefited from and contribute to it, so I'd like to thank the community!
>
> We're not writing new stuff in MR and instead use alternatives (like Spark) so I think this would make sense.
>
> +1
>
> Cosmin
>
> ________________________________________
> From: Konstantin Boudnik <co...@apache.org>
> Sent: Tuesday, March 15, 2016 12:33 PM
> To: dev@mrunit.apache.org
> Subject: [DISCUSS] Moving MRUnit to attic
>
> Hello.
>
> I want to start the discussion about the near future of the project. Here's my
> line of thinking. The MapReduce paradigm is clearly dying out and is getting
> replaced more and more either by new MapReduce-like platforms (think Spark) or
> by direct streaming processing and in-memory systems, completely going around
> the batch.
>
> MRUnit is definitely a good project to help developers to deal with MR
> realities before the code hits the cluster. But evidently there isn't much
> activity going on as dev@ list traffic (literally <10 JIRA messages/month) [1]
> and line of releases indicate. According to reporter.apache.org the last
> release had happened on May 2012, but I found some email evidence that 1.1.0
> was out in June 2014 [2](?)
>
> Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame in
> it. If anything, it is a part of the natural cycle of any software project.
> And most important, the code base will not go anywhere and can be resurrected
> any time if there's a community to work on it.
>
> Thoughts?
> Cos
>
> 1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
> 2. http://is.gd/mgFXsC
> 3. http://attic.apache.org/
>
Re: [DISCUSS] Moving MRUnit to attic
Posted by Cosmin Lehene <cl...@adobe.com>.
MRUnit has been a useful tool.
I'm glad to have benefited from and contribute to it, so I'd like to thank the community!
We're not writing new stuff in MR and instead use alternatives (like Spark) so I think this would make sense.
+1
Cosmin
________________________________________
From: Konstantin Boudnik <co...@apache.org>
Sent: Tuesday, March 15, 2016 12:33 PM
To: dev@mrunit.apache.org
Subject: [DISCUSS] Moving MRUnit to attic
Hello.
I want to start the discussion about the near future of the project. Here's my
line of thinking. The MapReduce paradigm is clearly dying out and is getting
replaced more and more either by new MapReduce-like platforms (think Spark) or
by direct streaming processing and in-memory systems, completely going around
the batch.
MRUnit is definitely a good project to help developers to deal with MR
realities before the code hits the cluster. But evidently there isn't much
activity going on as dev@ list traffic (literally <10 JIRA messages/month) [1]
and line of releases indicate. According to reporter.apache.org the last
release had happened on May 2012, but I found some email evidence that 1.1.0
was out in June 2014 [2](?)
Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame in
it. If anything, it is a part of the natural cycle of any software project.
And most important, the code base will not go anywhere and can be resurrected
any time if there's a community to work on it.
Thoughts?
Cos
1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
2. http://is.gd/mgFXsC
3. http://attic.apache.org/
Re: [DISCUSS] Moving MRUnit to attic
Posted by Konstantin Boudnik <co...@apache.org>.
The thread has been open for almost two weeks now, and as far I can see the
consensus is pretty much there. I will start a new thread to [VOTE] on the
proposed board resolution to move the project to the Attic.
Thanks
Cos
On Tue, Mar 15, 2016 at 12:33PM, Konstantin Boudnik wrote:
> Hello.
>
> I want to start the discussion about the near future of the project. Here's my
> line of thinking. The MapReduce paradigm is clearly dying out and is getting
> replaced more and more either by new MapReduce-like platforms (think Spark) or
> by direct streaming processing and in-memory systems, completely going around
> the batch.
>
> MRUnit is definitely a good project to help developers to deal with MR
> realities before the code hits the cluster. But evidently there isn't much
> activity going on as dev@ list traffic (literally <10 JIRA messages/month) [1]
> and line of releases indicate. According to reporter.apache.org the last
> release had happened on May 2012, but I found some email evidence that 1.1.0
> was out in June 2014 [2](?)
>
> Hence, is it a time perhaps to move MRUnit to attic [3]? There's no shame in
> it. If anything, it is a part of the natural cycle of any software project.
> And most important, the code base will not go anywhere and can be resurrected
> any time if there's a community to work on it.
>
> Thoughts?
> Cos
>
> 1. https://mail-archives.apache.org/mod_mbox/mrunit-dev/
> 2. http://is.gd/mgFXsC
> 3. http://attic.apache.org/
>