[Gluster-users] Gluster quotas, NFS quotas, brick quotas, quota-tools

Varun Shastry vshastry at redhat.com
Mon Mar 11 12:30:22 UTC 2013

Hi Leonardo,

Currently we don't have any method to import any back-end FS's quotas 
into gluster.

On Saturday 02 March 2013 03:26 AM, Leonardo Boiko wrote:
> Hi,
> I'd like to try to migrate our NFS-based network to gluster, for replication.
> We currently use an ext4 filesystem with quota enabled.  We have a
> couple thousand users (over LDAP), and their quotas vary wildly.  We
> manage them with edquota(8) on the NFS server.  There are many
> services which mount the NFS filesystem, and parse the output of
> /usr/bin/quota (a webmail, two mail servers, various webpages, etc.)
> I read the chapter on quotas in the admin guide, but I kind of don't
> understand yet how does Gluster directory quota relate to the usual
> quota stuff. Questions:
>   - Is there any way to make Gluster use the ext4 disk quotas, or at
> least import them?
No, its not possible (at least) as of now. By the way using ext4 
back-end and creating gluster volumes on existing data node is not 
***recommended**. *
> - I don't think edquota will work anymore? 
Gluster quota doesn't impair back-end FS quota behaviour, but gluster 
knows nothing about that, so you can't have a volume wide quota 
> - If I mount a gluster filesystem with type=glusterfs, will 
> /usr/bin/quota be able to read quotas from it?
No, only way to manage gluster quota is from glusterd (gluster 
management daemon) commands.
>   - If not, can I otherwise read my quotas from the client?
>   - If I mount a gluster filesystem using NFS, can the client read
> quotas with /usr/bin/quota as usual?
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20130311/f7b43c8f/attachment.html>

More information about the Gluster-users mailing list