raid1 - optimization

raid1 - optimization

[Roberto Spadim]

hi guys, i another thread i read about SNIA DDF Technical Position v2.0
could we implement RAID-1E, for sequencial/random read speedup?
http://www.snia.org/tech_activities/standards/curr_standards/ddf/SNIA_DDF_Technical_Position_v2.0.pdf
after that, could we implement some tweaks on read balance (there's
another thread just for it)

-- 
Roberto Spadim
Spadim Technology / SPAEmpresarial

Re: raid1 - optimization

[John Robinson]

On 03/02/2011 14:37, Roberto Spadim wrote:
> hi guys, i another thread i read about SNIA DDF Technical Position v2.0
> could we implement RAID-1E, for sequencial/random read speedup?
> http://www.snia.org/tech_activities/standards/curr_standards/ddf/SNIA_DDF_Technical_Position_v2.0.pdf
> after that, could we implement some tweaks on read balance (there's
> another thread just for it)

Those RAID-1E layouts look to me like what's implemented in md RAID as 
RAID-10 with near and offset layouts.

As discussed frequently here, md RAID-10 also has a far layout which can 
provide slightly faster than RAID-0 throughput for sequential reading 
with slightly slower than RAID-1 throughput for sequential writing. 
Please see some of the benchmarks that have also been referred to on 
this list in the last couple of days.

Cheers,

John.

Re: raid1 - optimization

[Roberto Spadim]

nice =] i will test it (i think that´s raid-10 with offset layout is
raid 1e on snia document)
the second point where could i find developers to change a bit current
read_balance of raid1 and raid10?

2011/2/3 John Robinson <john.robinson@anonymous.org.uk>:
> On 03/02/2011 14:37, Roberto Spadim wrote:
>>
>> hi guys, i another thread i read about SNIA DDF Technical Position v2.0
>> could we implement RAID-1E, for sequencial/random read speedup?
>>
>> http://www.snia.org/tech_activities/standards/curr_standards/ddf/SNIA_DDF_Technical_Position_v2.0.pdf
>> after that, could we implement some tweaks on read balance (there's
>> another thread just for it)
>
> Those RAID-1E layouts look to me like what's implemented in md RAID as
> RAID-10 with near and offset layouts.
>
> As discussed frequently here, md RAID-10 also has a far layout which can
> provide slightly faster than RAID-0 throughput for sequential reading with
> slightly slower than RAID-1 throughput for sequential writing. Please see
> some of the benchmarks that have also been referred to on this list in the
> last couple of days.
>
> Cheers,
>
> John.
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>



-- 
Roberto Spadim
Spadim Technology / SPAEmpresarial
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: raid1 - optimization

[Keld Jørn Simonsen]

On Thu, Feb 03, 2011 at 02:56:49PM +0000, John Robinson wrote:
> On 03/02/2011 14:37, Roberto Spadim wrote:
> >hi guys, i another thread i read about SNIA DDF Technical Position v2.0
> >could we implement RAID-1E, for sequencial/random read speedup?
> >http://www.snia.org/tech_activities/standards/curr_standards/ddf/SNIA_DDF_Technical_Position_v2.0.pdf
> >after that, could we implement some tweaks on read balance (there's
> >another thread just for it)
> 
> Those RAID-1E layouts look to me like what's implemented in md RAID as 
> RAID-10 with near and offset layouts.
> 
> As discussed frequently here, md RAID-10 also has a far layout which can 
> provide slightly faster than RAID-0 throughput for sequential reading 
> with slightly slower than RAID-1 throughput for sequential writing. 
> Please see some of the benchmarks that have also been referred to on 
> this list in the last couple of days.

I have written SNIA over a year ago on raid10,f2 to have it
to be a standardized layout too, but without any response.

Best regards
Keld