Missing GlusterFS in the datastore?

Hello,

I want to setup a test ONE 5.0 but I can’t find GlusterFS in the storage documentation.

Is the support of GlusterFS removed?

Regards.

The native access through libvirt is still supported but not recommended,
the preferred way is to access gluster through the shared/qcow2 drivers.
As I said the code is still there and functional if you have gluster
datastores

Cheers

I’m in the same boat. We are looking at 5.0 with the GlusterFS storage. When you say use the qcow2 drivers, is there anything special that must be done when creating the image other than selecting the driver as qcow2 and creating it on a “Filesystem - qcow2 mode” shared filesystem pointing to the GlusterFS mount?

Daryl Lee forum@opennebula.org writes:

I’m in the same boat. We are looking at 5.0 with the GlusterFS
storage. When you say use the qcow2 drivers, is there anything
special that must be done when creating the image other than selecting
the driver as qcow2 and creating it on a “Filesystem - qcow2 mode”
shared filesystem pointing to the GlusterFS mount?

That’s what I did:

  • create my GlusterFS volume one-datastores

  • mount the GlusterFS volume under /var/lib/one/datastores

  • use qcow2 images with appropriate TM_MAD

Regards.
Daniel Dehennin
Récupérer ma clef GPG: gpg --recv-keys 0xCC1E9E5B7A6FE2DF
Fingerprint: 3E69 014E 5C23 50E8 9ED6 2AAD CC1E 9E5B 7A6F E2DF

Why this is not recommended ? There are many people that are using gluster :frowning: , gluster fuse mount is slower than using libgfsapi…

I came here to ask this. Since I noticed this thread I had been looking around for some new support directly related to the qcow2 driver that avoids the traditional FUSE method but was unable to find it. This is a good question, curious if i’m missing something.

Unfortunately we don’t have infinite resources and we can not fully support all technologies under the sun.

  • In the latest survey not that many people said that used glusterfs
  • OpenNebula GlusterFS driver was lacking and didn’t support some features like system/disk snapshotting
  • GlusterFS can be used as an alternative to NFS and qcow2 drivers work perfectly fine with it

Moreover, as Ruben said, the code to support GlusterFS was not removed from the repository, only from the documentation.

Thanks a lot for your reply Javi. Moreover, I would like to ask if we should start thinking of changing gluster environments to CEPH or simular infrastractures/distributed filesystems. I am afraid that in next releases you could remove GlusterFS support and it will be “hell” situation for “us” or mainly for our newly deployed infrastructures. We will probably switch to CEPH if your answer will be like “do not use GlusterFS with OpenNebula anymore anymore if you want to be with us and stable”.

This is so true, but I think GlusterFS is growing and it becomes usable (in means of deploying GlusterFS in production environments) even more than 1 year before, snapshots are relatively new feature maybe it’s because of that. I hope GlusterFS will become more and more stable this year.

Hi everyone!

Support of this FS critical for us too.
We started our work with ONE because it had a native GlusterFS driver.
But now we obliged to use Datastore through FUSE.
It’s really awful moment.

But as I understood - all thing in the hands of developers.
So glusterfs/ceph/lizard should be in ONE as it was.

Of course this is just opinion one of customers…
But maybe we can provide some help or provide couple developers?

1 Like