[Gluster-users] [Gluster3.2 at Grid5000] 128 nodes failure and rr scheduler question

[Amar Tumballi]

Hi Francois,

Answers inline.

On Wed, Jun 8, 2011 at 6:10 PM, Francois THIEBOLT <thiebolt at irit.fr> wrote:

> Hello,
>
> I'm driving some experiments on grid'5000 with GlusterFS 3.2 and, as a
> first point, i've been unable to start a volume featuring 128bricks (64 ok)
>
> This looks similar to the bug http://bugs.gluster.com/show_bug.cgi?id=2941 The
fix should be available with 3.2.1 release, which should be out very soon.

Also we are working on scalability of 'glusterd', the glusterfs management
daemon, after which it should work fine. One work around for now is, create
a volume with 64 bricks and do the 'add-brick' of another 64 brick. It
should work fine.



> Then, due to the round-robin scheduler, as the number of nodes increase
> (every node is also a brick), the performance of an application on an
> individual node decrease!
> So my question is : how to STOP the round-robin distribution of files over
> the bricks within a volume ?
>

There is no 'Scheduler' in picture here with GlusterFS 3.2.x (for that
matter from 3.0.x releases), hence there is no option to stop the scheduler.

Regards,
Amar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20110610/f2bccf27/attachment.html>