drop i386 support

drop i386 support

[kefu chai]

Hi Cephers,

I am proposing drop the support of i386. as we don't compile Ceph with
any i386 gitbuilder now[1] and hence don't test the i386 builds on
sepia on a regular basis. Also, based on the assumption that people
don't use i386 in production, I think we can drop it from the minimum
hardware document[2]?

And we won't explicitly disable the i386 build in code if we decide to
drop the i386 support, as we always try to be portable if possible.
But just don't claim the i386 as the officially supported arch
anymore.

What do you think?

---
[1] http://ceph.com/gitbuilder.cgi
[2] http://docs.ceph.com/docs/master/start/hardware-recommendations/#minimum-hardware-recommendations


-- 
Regards
Kefu Chai

Re: drop i386 support

[Piotr Dałek]

On Mon, Jul 11, 2016 at 02:57:12PM +0800, kefu chai wrote:
> Hi Cephers,
> 
> I am proposing drop the support of i386. as we don't compile Ceph with
> any i386 gitbuilder now[1] and hence don't test the i386 builds on
> sepia on a regular basis. Also, based on the assumption that people
> don't use i386 in production, I think we can drop it from the minimum
> hardware document[2]?
> 
> And we won't explicitly disable the i386 build in code if we decide to
> drop the i386 support, as we always try to be portable if possible.
> But just don't claim the i386 as the officially supported arch
> anymore.
> 
> What do you think?

+1 from my side. Regular users are on amd64, and those willing to run on
small devices are using ARM anyway.

-- 
Piotr Dałek
branch@predictor.org.pl
http://blog.predictor.org.pl
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: drop i386 support

[Wido den Hollander]

> Op 11 juli 2016 om 9:32 schreef Piotr Dałek <branch@predictor.org.pl>:
> 
> 
> On Mon, Jul 11, 2016 at 02:57:12PM +0800, kefu chai wrote:
> > Hi Cephers,
> > 
> > I am proposing drop the support of i386. as we don't compile Ceph with
> > any i386 gitbuilder now[1] and hence don't test the i386 builds on
> > sepia on a regular basis. Also, based on the assumption that people
> > don't use i386 in production, I think we can drop it from the minimum
> > hardware document[2]?
> > 
> > And we won't explicitly disable the i386 build in code if we decide to
> > drop the i386 support, as we always try to be portable if possible.
> > But just don't claim the i386 as the officially supported arch
> > anymore.
> > 
> > What do you think?
> 
> +1 from my side. Regular users are on amd64, and those willing to run on
> small devices are using ARM anyway.
> 

Indeed. In my experience I have never seen a cluster run on i386.

+1 from my side to drop i386.

> -- 
> Piotr Dałek
> branch@predictor.org.pl
> http://blog.predictor.org.pl
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: drop i386 support

[Sage Weil]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 1258 bytes --]

On Mon, 11 Jul 2016, Wido den Hollander wrote:
> > Op 11 juli 2016 om 9:32 schreef Piotr Dałek <branch@predictor.org.pl>:
> > On Mon, Jul 11, 2016 at 02:57:12PM +0800, kefu chai wrote:
> > > Hi Cephers,
> > > 
> > > I am proposing drop the support of i386. as we don't compile Ceph with
> > > any i386 gitbuilder now[1] and hence don't test the i386 builds on
> > > sepia on a regular basis. Also, based on the assumption that people
> > > don't use i386 in production, I think we can drop it from the minimum
> > > hardware document[2]?
> > > 
> > > And we won't explicitly disable the i386 build in code if we decide to
> > > drop the i386 support, as we always try to be portable if possible.
> > > But just don't claim the i386 as the officially supported arch
> > > anymore.
> > > 
> > > What do you think?
> > 
> > +1 from my side. Regular users are on amd64, and those willing to run on
> > small devices are using ARM anyway.
> > 
> 
> Indeed. In my experience I have never seen a cluster run on i386.
> 
> +1 from my side to drop i386.

Yep.

The only value is the i386 builder, which serves as a canary for 32-bit 
build issues, and is a more convenient that building on 32-bit arm 
hardware.  So +.5 from me to keeping the builder around...

sage

Re: drop i386 support

[Alfredo Deza]

On Mon, Jul 11, 2016 at 8:58 AM, Sage Weil <sage@newdream.net> wrote:
> On Mon, 11 Jul 2016, Wido den Hollander wrote:
>> > Op 11 juli 2016 om 9:32 schreef Piotr Dałek <branch@predictor.org.pl>:
>> > On Mon, Jul 11, 2016 at 02:57:12PM +0800, kefu chai wrote:
>> > > Hi Cephers,
>> > >
>> > > I am proposing drop the support of i386. as we don't compile Ceph with
>> > > any i386 gitbuilder now[1] and hence don't test the i386 builds on
>> > > sepia on a regular basis. Also, based on the assumption that people
>> > > don't use i386 in production, I think we can drop it from the minimum
>> > > hardware document[2]?
>> > >
>> > > And we won't explicitly disable the i386 build in code if we decide to
>> > > drop the i386 support, as we always try to be portable if possible.
>> > > But just don't claim the i386 as the officially supported arch
>> > > anymore.
>> > >
>> > > What do you think?
>> >
>> > +1 from my side. Regular users are on amd64, and those willing to run on
>> > small devices are using ARM anyway.
>> >
>>
>> Indeed. In my experience I have never seen a cluster run on i386.
>>
>> +1 from my side to drop i386.
>
> Yep.
>
> The only value is the i386 builder, which serves as a canary for 32-bit
> build issues, and is a more convenient that building on 32-bit arm
> hardware.  So +.5 from me to keeping the builder around...

We also don't build i386 releases anymore (at least for a couple of years now)


>
> sage
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: drop i386 support

[kefu chai]

On Mon, Jul 11, 2016 at 9:42 PM, Alfredo Deza <adeza@redhat.com> wrote:
> On Mon, Jul 11, 2016 at 8:58 AM, Sage Weil <sage@newdream.net> wrote:
>> On Mon, 11 Jul 2016, Wido den Hollander wrote:
>>> > Op 11 juli 2016 om 9:32 schreef Piotr Dałek <branch@predictor.org.pl>:
>>> > On Mon, Jul 11, 2016 at 02:57:12PM +0800, kefu chai wrote:
>>> > > Hi Cephers,
>>> > >
>>> > > I am proposing drop the support of i386. as we don't compile Ceph with
>>> > > any i386 gitbuilder now[1] and hence don't test the i386 builds on
>>> > > sepia on a regular basis. Also, based on the assumption that people
>>> > > don't use i386 in production, I think we can drop it from the minimum
>>> > > hardware document[2]?
>>> > >
>>> > > And we won't explicitly disable the i386 build in code if we decide to
>>> > > drop the i386 support, as we always try to be portable if possible.
>>> > > But just don't claim the i386 as the officially supported arch
>>> > > anymore.
>>> > >
>>> > > What do you think?
>>> >
>>> > +1 from my side. Regular users are on amd64, and those willing to run on
>>> > small devices are using ARM anyway.
>>> >
>>>
>>> Indeed. In my experience I have never seen a cluster run on i386.
>>>
>>> +1 from my side to drop i386.
>>
>> Yep.
>>
>> The only value is the i386 builder, which serves as a canary for 32-bit
>> build issues, and is a more convenient that building on 32-bit arm
>> hardware.  So +.5 from me to keeping the builder around...
>
> We also don't build i386 releases anymore (at least for a couple of years now)


thanks a lot, guys! i am posting a PR[1] removing i386 from the
minimal hardware requirement doc.

---
[1] https://github.com/ceph/ceph/pull/10276


-- 
Regards
Kefu Chai
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html