I have a question for the developers: Why use gluster? Why not Pacemaker or something with better performance stats?

Just curious.

Eric Evans

Digital Data Services LLC.

304.660.9080

From: Joop <jvdwege@xs4all.nl>
Sent: Monday, April 13, 2020 9:43 AM
To: Satheesaran Sundaramoorthi <sasundar@redhat.com>; users@ovirt.org
Subject: [ovirt-users] Re: oVirt 4.4.0 Beta release refresh is now available for testing

On 13-4-2020 09:45, Satheesaran Sundaramoorthi wrote:

On Thu, Apr 9, 2020 at 3:37 PM Sandro Bonazzola <sbonazzo@redhat.com> wrote:

Does the hyperconverged installation now work with glusterfs?
I tested alpha/beta1/2 and the latter won't get past the storage step.

Not yet. There was the issue with glusterfs storage domain and storage domain blocksize probe check.

This is addressed with a new ioprocess-1.4.1 package for the bug[1].

We are testing currently with this fix and it works well. But hitting yet another issue, where Hosted Engine deployment

fails at the very last stage in the deployment. Gluster developer is looking in to it, as I respond here.

Soon we will have a working build and I will let you once these issues are settled.

[1] - https://bugzilla.redhat.com/show_bug.cgi?id=1820283

Thanks,

sas

Thanks for the message. Looking through the logs I saw references to the blocksize but since that is contained in the ansible scripts which get downloaded to the host I gave up finding the source versions to try to see it I could get it going beyond that failing step.

Anyway, thanks for informing me,

Regards,

Joop