You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by Steve Loughran <st...@cloudera.com.INVALID> on 2020/01/06 11:38:41 UTC
[DISCUSS] drop mapreduce-pipes
It has been many years since anyone did any work on mapreduce-pipes and
given the large number of hadoop-cluster streaming frameworks in the ASF
alone, probably the same number of years since anyone actually launched a
pipe job other than during testing.
As such, it's a background noise maintenance problem and extra delays in
test runs.
What do people think about deleting it?
-Steve
Re: [DISCUSS] drop mapreduce-pipes
Posted by Masatake Iwasaki <iw...@oss.nttdata.co.jp>.
I'm +1 too.
Marking it as deprecated on 3.3.0 and dropping on 3.4.0 would be an option.
-Pnative build failure on RHEL/CentOS 8 is on-going issue of hadoop-pipes.
I will appreciate review of HADOOP-16739 if it takes some time to drop
pipes.
Thanks,
Masatake Iwasaki
On 2020/01/07 1:26, Esteban Gutierrez wrote:
> As long as there is a proper announcement and deprecation path I'm +1. I
> know some third party products relied on hadoop pipes as a workaround to
> run parallelized version of their existing binaries, so assuming some are
> still being used people should be aware of this proposal.
>
> Thanks!
> esteban.
>
>
> --
> Cloudera, Inc.
>
>
>
> On Mon, Jan 6, 2020 at 8:52 AM Brahma Reddy Battula <br...@apache.org>
> wrote:
>
>> +1 on deleting this .
>>
>>
>> On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <stevel@cloudera.com.invalid
>> wrote:
>>
>>> It has been many years since anyone did any work on mapreduce-pipes and
>>> given the large number of hadoop-cluster streaming frameworks in the ASF
>>> alone, probably the same number of years since anyone actually launched a
>>> pipe job other than during testing.
>>>
>>> As such, it's a background noise maintenance problem and extra delays in
>>> test runs.
>>>
>>> What do people think about deleting it?
>>>
>>> -Steve
>>>
>> --
>>
>>
>>
>> --Brahma Reddy Battula
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-dev-unsubscribe@hadoop.apache.org
For additional commands, e-mail: mapreduce-dev-help@hadoop.apache.org
Re: [DISCUSS] drop mapreduce-pipes
Posted by Masatake Iwasaki <iw...@oss.nttdata.co.jp>.
I'm +1 too.
Marking it as deprecated on 3.3.0 and dropping on 3.4.0 would be an option.
-Pnative build failure on RHEL/CentOS 8 is on-going issue of hadoop-pipes.
I will appreciate review of HADOOP-16739 if it takes some time to drop
pipes.
Thanks,
Masatake Iwasaki
On 2020/01/07 1:26, Esteban Gutierrez wrote:
> As long as there is a proper announcement and deprecation path I'm +1. I
> know some third party products relied on hadoop pipes as a workaround to
> run parallelized version of their existing binaries, so assuming some are
> still being used people should be aware of this proposal.
>
> Thanks!
> esteban.
>
>
> --
> Cloudera, Inc.
>
>
>
> On Mon, Jan 6, 2020 at 8:52 AM Brahma Reddy Battula <br...@apache.org>
> wrote:
>
>> +1 on deleting this .
>>
>>
>> On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <stevel@cloudera.com.invalid
>> wrote:
>>
>>> It has been many years since anyone did any work on mapreduce-pipes and
>>> given the large number of hadoop-cluster streaming frameworks in the ASF
>>> alone, probably the same number of years since anyone actually launched a
>>> pipe job other than during testing.
>>>
>>> As such, it's a background noise maintenance problem and extra delays in
>>> test runs.
>>>
>>> What do people think about deleting it?
>>>
>>> -Steve
>>>
>> --
>>
>>
>>
>> --Brahma Reddy Battula
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: common-dev-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-dev-help@hadoop.apache.org
Re: [DISCUSS] drop mapreduce-pipes
Posted by Esteban Gutierrez <es...@cloudera.com.INVALID>.
As long as there is a proper announcement and deprecation path I'm +1. I
know some third party products relied on hadoop pipes as a workaround to
run parallelized version of their existing binaries, so assuming some are
still being used people should be aware of this proposal.
Thanks!
esteban.
--
Cloudera, Inc.
On Mon, Jan 6, 2020 at 8:52 AM Brahma Reddy Battula <br...@apache.org>
wrote:
> +1 on deleting this .
>
>
> On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <stevel@cloudera.com.invalid
> >
> wrote:
>
> > It has been many years since anyone did any work on mapreduce-pipes and
> > given the large number of hadoop-cluster streaming frameworks in the ASF
> > alone, probably the same number of years since anyone actually launched a
> > pipe job other than during testing.
> >
> > As such, it's a background noise maintenance problem and extra delays in
> > test runs.
> >
> > What do people think about deleting it?
> >
> > -Steve
> >
> --
>
>
>
> --Brahma Reddy Battula
>
Re: [DISCUSS] drop mapreduce-pipes
Posted by Esteban Gutierrez <es...@cloudera.com.INVALID>.
As long as there is a proper announcement and deprecation path I'm +1. I
know some third party products relied on hadoop pipes as a workaround to
run parallelized version of their existing binaries, so assuming some are
still being used people should be aware of this proposal.
Thanks!
esteban.
--
Cloudera, Inc.
On Mon, Jan 6, 2020 at 8:52 AM Brahma Reddy Battula <br...@apache.org>
wrote:
> +1 on deleting this .
>
>
> On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <stevel@cloudera.com.invalid
> >
> wrote:
>
> > It has been many years since anyone did any work on mapreduce-pipes and
> > given the large number of hadoop-cluster streaming frameworks in the ASF
> > alone, probably the same number of years since anyone actually launched a
> > pipe job other than during testing.
> >
> > As such, it's a background noise maintenance problem and extra delays in
> > test runs.
> >
> > What do people think about deleting it?
> >
> > -Steve
> >
> --
>
>
>
> --Brahma Reddy Battula
>
Re: [DISCUSS] drop mapreduce-pipes
Posted by Brahma Reddy Battula <br...@apache.org>.
+1 on deleting this .
On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <st...@cloudera.com.invalid>
wrote:
> It has been many years since anyone did any work on mapreduce-pipes and
> given the large number of hadoop-cluster streaming frameworks in the ASF
> alone, probably the same number of years since anyone actually launched a
> pipe job other than during testing.
>
> As such, it's a background noise maintenance problem and extra delays in
> test runs.
>
> What do people think about deleting it?
>
> -Steve
>
--
--Brahma Reddy Battula
Re: [DISCUSS] drop mapreduce-pipes
Posted by Brahma Reddy Battula <br...@apache.org>.
+1 on deleting this .
On Mon, 6 Jan 2020 at 7:39 PM, Steve Loughran <st...@cloudera.com.invalid>
wrote:
> It has been many years since anyone did any work on mapreduce-pipes and
> given the large number of hadoop-cluster streaming frameworks in the ASF
> alone, probably the same number of years since anyone actually launched a
> pipe job other than during testing.
>
> As such, it's a background noise maintenance problem and extra delays in
> test runs.
>
> What do people think about deleting it?
>
> -Steve
>
--
--Brahma Reddy Battula