[Openstack-operators] Advice about distributed FS for glance/nova

Tim Bell Tim.Bell at cern.ch
Tue Dec 17 08:05:01 UTC 2013


At CERN, we use ceph for glance and cinder with Scientific Linux 6. You’ll need a patched QEMU from Inktank to support the cinder backing store.

The user survey (http://www.openstack.org/blog/2013/11/openstack-user-survey-october-2013/) has the most commonly deployed options for backing store with NFS, Ceph, GlusterFS as the open source solutions and Netapp or EMC for the commercial ones.

Tim

From: Nick Maslov [mailto:azpekt at gmail.com]
Sent: 17 December 2013 08:17
To: Alvise Dorigo; Razique Mahroua
Cc: openstack-operators at lists.openstack.org
Subject: Re: [Openstack-operators] Advice about distributed FS for glance/nova

Hi,

Using Ceph here as well. It installs easily, works nicely, integrates with OpenStack modules (Glance, Cinder, Swift) as well.

Cheers,
NM

--
Nick Maslov
Sent with Airmail


On November 28, 2013 at 6:43:12 PM, Razique Mahroua (razique.mahroua at gmail.com<mailto://razique.mahroua@gmail.com>) wrote:

Hi :)
We discussed a lot about that. First and direct answer: you won't find "the best".

It all boils down to:
1. What you are intend to do
2. What is your current hardware
3. What kind of load you at least expect.

Here is a first thread:
http://www.gossamer-threads.com/lists/openstack/dev/24231

and another interesting one:
http://www.gossamer-threads.com/lists/openstack/dev/24231

in a nutshell, I myself tested many clustered or shared FS, and they all have their pros and cons.

In my top list, I ended up having:
- MooseFS
- Ceph (but not CephFS)
- ISCSI

only because they happened to work well with my environments, they "did the job"

Let us know if you have any specific questions regarding these technologies

Razique
On 28 Nov 2013, at 0:44, Alvise Dorigo wrote:

Hello,
I'm setting up an openstack-based cloud IaaS. Everything is going to be in HA mode.
In particular I'm wondering which would be the best choice for the backend FS to support the storage for glance and nova (i.e. /var/lib/glance/images and /var/lib/nova/instances).

I've read around; there's a number of FS and a lot discussions. The most cited and used seem to be (not only for cloud IaaS):
GlusterFS
LustreFS
CEPH
GPFS
MooseFS
FhGFS
CODA
NFS
AFS

My requirements are:
- extendability (new disk array transparently added without any outage)
- high performance on big files (glance and nova usually manage big files)
- POSIX / mount as a normal filesystem with unique naming space
- easy to install/manage (and supported on RHEL 6.x)

Any advice or report on your experience would be very appreciated.

Regards,

Alvise_______________________________________________



OpenStack-operators mailing list
OpenStack-operators at lists.openstack.org<mailto:OpenStack-operators at lists.openstack.org>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

________________________________
- signature.asc, 548 bytes
_______________________________________________
OpenStack-operators mailing list
OpenStack-operators at lists.openstack.org<mailto:OpenStack-operators at lists.openstack.org>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20131217/a2100db7/attachment.html>


More information about the OpenStack-operators mailing list