[Gluster-users] Why does glusterfs has nfs stuff on the server

[Shehjar Tikoo]

Mohit Anchlia wrote:
> Can you please explain what do you mean by "All Gluster volumes are
> exported through nfs"? I thought gluster just uses fuse on the server
> and then client can decide to use nfs or not.
That is correct. The client can decide to use nfs BUT nfs needs to be 
enabled if the user has to mount using nfs. This is enabled by default and 
that is why you see NFS related process started automatically. If you dont 
want it, see my earlier reply.

-Shehjar


> 
> But what I am seeing is that after installing gluster on the "server"
> and then do a ps on gluster process it shows "nfs-server.vol, nfs.log
> etc." Why?
> 
> Thanks for your response
> 
> On Mon, Mar 21, 2011 at 10:48 PM, Shehjar Tikoo <shehjart at gluster.com> wrote:
>> All Gluster volumes are exported through nfs by default. To disable nfs on
>> 3.1.3 release, use the nfs.disable command line option. For more info on
>> this, please see the release notes.
>>
>> Mohit Anchlia wrote:
>>> When I installed gluster and do a "ps" on the process I see:
>>>
>>> /usr/sbin/glusterfs -f /etc/glusterd/nfs/nfs-server.vol -p
>>> /etc/glusterd/nfs/run/nfs.pid -l /var/log/glusterfs/nfs.log"
>>>
>>> My question is why did glusterfs use nfs-server.vol, nfs.pid and
>>> nfs.log instead of using some generic name. This is confusing and
>>> makes me think it's using nfs somehow on the server even though that
>>> doesn't look like it.
>>>
>>> We use direct attached storage, not NFS. This seems to come with
>>> default installation of gluster. Is this just a mistake in how scripts
>>> were named or is there more to it?
>>> _______________________________________________
>>> Gluster-users mailing list
>>> Gluster-users at gluster.org
>>> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>>
>>
>>