[one-users] experiences with distributed FS?

[Michael Berlin]

Hi,

I'm one of the developers of the distributed file system XtreemFS. I 
just noticed this thread since one of the posters followed us on 
Twitter. So I just joint the mailing list.

Internally we also run a OpenNebula cluster and use XtreemFS as shared 
file system there. Since XtreemFS is POSIX compatible, you can use the 
tm_nfs and just point OpenNebula to the mounted POSIX volume.

A colleague of mine did setup the installation which is based on 
OpenNebula 2.2. So I'm not fully aware of the details. He told me that 
he has also pointed the variable VM_DIR (in $ONE_LOCATION/etc/oned.conf) 
to a XtreemFS volume, so our KVM-based installation does support live 
migration. So far we haven't seen any problems with using XtreemFS in 
OpenNebula, otherwise we would have fixed it ;-)

Regarding the performance: We did not do any measurements so far. Any 
numbers or suggestions how to benchmark different distributed file 
systems are welcome.

In XtreemFS the metadata of a volume is stored on one metadata server. 
If you like, you can set it up replicated and then your file system will 
have no single point of failure.

Regarding support: Although we cannot offer commercial support, we 
provide support through our mailing list and are always eager to help.

Best regards,
Michael


On 02/09/2012 09:52 AM, Gian Uberto Lauri wrote:
>>>>>> "HE" == Hans-Joachim Ehlers<HansJoachim.Ehlers at eumetsat.int>
>>>>>> writes:
>
> HE>  Hi Joao,
>>> We're looking into the possibility of using a distributed FS, like
>>> glusterfs or similar solution for the image repository
>
> HE>  We are looking forward to use GPFS as a clustered FS for our
> HE>  upcoming KVM&  OpenNebula solution since we are already a happy
> HE>  GPFS user.
>
> Beware  of the  clustering  software, be  sure to  use  a very  recent
> release (i.e. the on in Debian Stable).
>
> I tried it with an older, broken version of the clustering software and
> initially it worked wonderfully well... Until the cluster software began
> to malfunction and kick the servers out.
>
> But we plan  to use something like  this in the future  (since we have
> sysadmin much better than me available :) ).
>
> --
> ing. Gian Uberto Lauri
> Ricercatore / Reasearcher
> Laboratorio Ricerca e Sviluppo / Research&  Development Lab.
> Area Calcolo Distribuito / Distributed Computation Area
>
> GianUberto.Lauri at eng.it
>
> Engineering Ingegneria Informatica spa
> Corso Stati Uniti 23/C, 35127 Padova (PD)
>
> Tel. +39-049.8283.571         | main(){printf(&unix["\021%six\012\0"],
> Fax  +39-049.8283.569		  |    (unix)["have"]+"fun"-0x60);}
> Skype: gian.uberto.lauri	  |          David Korn, AT&T Bell Labs
> http://www.eng.it 			  |          ioccc best One Liner, 1987
> _______________________________________________
> Users mailing list
> Users at lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org