Btrfs devel - Feb 2010 - does btrfs have RAID I/O throughput (un)limiting sysctls, similar to md?

i''ve a 4 drive array connected via a PCIe SATA card.

per OS (opensuse) default, md RAID I/O performance was being limited by,

  cat /proc/sys/dev/raid/speed_limit_min
    1000
  cat /proc/sys/dev/raid/speed_limit_max
    200000

changing,

  echo "dev.raid.speed_limit_min=100000" >> /etc/sysctl.conf
  echo "dev.raid.speed_limit_max=600000" >> /etc/sysctl.conf

enabled full/best I/O throughput.

does btrfs have a similar construct that I need to set/tweak for
maximum I/O throughput?
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Sat, Feb 6, 2010 at 12:37 AM, 0bo0 <0.bugs.only.0@gmail.com>
wrote:
> i''ve a 4 drive array connected via a PCIe SATA card.
>
> per OS (opensuse) default, md RAID I/O performance was being limited by,
>
>  cat /proc/sys/dev/raid/speed_limit_min
>    1000
>  cat /proc/sys/dev/raid/speed_limit_max
>    200000
>
> changing,
>
>  echo "dev.raid.speed_limit_min=100000" >> /etc/sysctl.conf
>  echo "dev.raid.speed_limit_max=600000" >> /etc/sysctl.conf
>
> enabled full/best I/O throughput.
These proc entries affect just array reconstruction, not general I/O
performance/throughput, so affect just an edge-case of applications
requiring maximum latency/minimum throughout guarantees.
-- 
Daniel J Blueman
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Sat, Feb 6, 2010 at 5:10 AM, Daniel J Blueman
<daniel.blueman@gmail.com> wrote:
> These proc entries affect just array reconstruction, not general I/O
> performance/throughput, so affect just an edge-case of applications
> requiring maximum latency/minimum throughout guarantees.
although i''d 1st seen the perf hit at the (re)construction stage, i
didn''t recognize that the sysctls were limited to that case.

so, iiuc, btrfs has no such issues?

thanks for clarifying!
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Am Samstag 06 Februar 2010 schrieben Sie:
> On Sat, Feb 6, 2010 at 5:10 AM, Daniel J Blueman
> 
> <daniel.blueman@gmail.com> wrote:
> > These proc entries affect just array reconstruction, not general I/O
> > performance/throughput, so affect just an edge-case of applications
> > requiring maximum latency/minimum throughout guarantees.
> 
> although i''d 1st seen the perf hit at the (re)construction stage,
i
> didn''t recognize that the sysctls were limited to that case.
Limiting rebuild throughput helps to have productively running 
applications getting their share of I/O during RAID rebuilds.
 
> so, iiuc, btrfs has no such issues?
Issues? I would call that a feature, not an issue.

-- 
Martin ''Helios'' Steigerwald - http://www.Lichtvoll.de
GPG: 03B0 0D6C 0040 0710 4AFA  B82F 991B EAAC A599 84C7