[Gluster-devel] guidance on size/characteristics of name space volume for cluster with ~750TB of data

[Amar S. Tumballi]

Hi JR,
 Currently namespace is used for holding global view of whole volume. The
only criteria of namespace would be requirement of inodes. if your namespace
can provide enough inode count for all your data, that should be sufficient.


(I use reiserfs for namespace, it handles small files (here 0byte) well, and
fast, and namespace data is not at all critical as its just used as
persistent cache).

Regards,
Amar


On Fri, May 30, 2008 at 7:42 AM, jrs <botemout at gmail.com> wrote:

> Greetings,
>
> In my testing of gluster I've just put my name space directory
> (which I'm replicating with afr) on /.
>
> Is this standard practice?  If not, are there any recommendations
> for how large the volume should (as a fraction of expected data
> size)?
>
> Thanks
> JR
>
>
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
Amar Tumballi
Gluster/GlusterFS Hacker
[bulde on #gluster/irc.gnu.org]
http://www.zresearch.com - Commoditizing Super Storage!