[Gluster-users] du time on bricks vs mounted glusterfs via fuse

[Gavin Henry]

Hi,

Completely forgot to say this is a mount on the same server.
On 6 May 2014 21:17, "Gavin Henry" <ghenry at surevoip.co.uk> wrote:

> Hi all,
>
> We use glusterfs with our SIP cluster for hosted SIP endpoints/phones.
> This is set up in replicated mode for lua scripts, voicemails (wav
> files), faxes (PDF's/tiffs) and call recordings (wav/mp3). We're
> trying to find more information on the time differences between a "du"
> on a brick, i.e. direct filesystem vs a "du" on a mounted export:
>
> box1 # time /usr/local/nagios/libexec/check_subdirectory_size
> /opt/surevoip/recordings/
> WARNING: blah size is 34126MB
>
> real 0m20.788s
> user 0m0.178s
> sys 0m0.559s
>
>
> box2 # time /usr/local/nagios/libexec/check_subdirectory_size
> /data/glusterfs/surevoip/brick1/brick/recordings/
> WARNING: blah size is 34195MB
>
> real 0m0.577s
> user 0m0.074s
> sys 0m0.161s
>
> Can anyone point me to the right documentation to read?
>
> Thanks,
>
> Gavin.
>
> --
> Kind Regards,
>
> Gavin Henry.
> http://www.surevoip.co.uk
>
> Did you see our API? http://www.surevoip.co.uk/api
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20140506/b8f5167b/attachment.html>