Hi Randy,

Openstack’s operations bases on the installed hypervisor. First, you need to know exactly your hypervisor that is supporting which types  of back-end storage.  After that, you can select one  back-end storage system corresponds to your system.

Cinder is just an front-end to help vms can connect to an back-end storage behind. Cinder supports so much driver to connect to back-end storage system such as: NFS, Ceph, GlusterFS, LVM ,…

If you have a back-end storage which cinder hasn’t supported yet driver to connect to, you can create a middle mount-point by using  NFS, ISCSI to help cinder can operate on them J.

Regards,

Khanh Nguyen

From: Randy S [mailto:sim.ple@live.nl]
Sent: Thursday, July 4, 2013 5:32 PM
To: community@lists.openstack.org
Subject: [openstack-community] usage of openstack information

Hi all,
 
I have been reading up on openstack as I am completely new to this. I have set 
up some test systems according to manuals found here and there.
Now, I have a question about the possible use of openstack storage.
I have seen that the normal provisioning of storage is meant to be for vm's 
created within openstack itself. This can then be block storage (cinder) or 
object storage.
 
Can somebody tell me if it is possible to use openstack to provision the same 
kind of storage to physical servers which are otherwise in no way connected to 
openstack or virtual machines created in systems like vmware?
 
I hope somebody can asnwer me this and maybe point me the way to test something 
like this.
 
Thanks in advance,
 
Rgds,
 
Randy