6:23 a.m.
You can reinstall your node - it will be assigned ryzen-based model for the cluster.Once
you import the storage domains, all VMs should be visible. Then you can edit each VM,
verify NIC, MAC address and change the CPU type of the VM to the one that is available in
the Cluster (respectfully the host).
Best Regards,Strahil Nikolov
On Sat, Mar 26, 2022 at 19:02, Nenhum_de_Nos<lojas(a)arroway.org> wrote: On Sat,
26 Mar 2022 06:37:39 +0000 (UTC)
Strahil Nikolov via Users <users(a)ovirt.org> wrote:
Can you summarize the situation, as it's hard to get an idea of
the
situation? On the Hypervisor: rpm -qa | grep ovirt | sortvirsh -c
qemu:///system?authfile=/etc/ovirt-hosted-engine/virsh_auth.conf
domcapabilities From the Engine UI:Cluster version and CPU type From
the HostedEngine VM:rpm -qa | grep ovirt | sort Best Regards,
Nikolov
Hi Strahil,
I know it was not the best explanation of the issue, I was kind
stressed as my VM's were offline for almost a week. Now things are
quite again.
I think this problem I went through is not quite common to ovirt users
as by what I could see here, people use server hardware, unlike me.
After the power outage I could not get my node back active on the
cluster (there is just one node here). And no matter how many times I
tried, the logs kept saying the CPU was not compatible. I am answering
here as I could not find this on my search through google and mail
lists, forums (just could not see the redhat forum, I don't have
account there).
My problem was that my machine got some kind of bios reset, at least it
disabled Virtualization Capabilities. And as I had not turned it off, I
never would suspect it was this.
In a desperate move, I installed ESXi on the machine and begun
migrating my VM's. When the first was able to run, he hypervisor told
me mu CPU was not VM capable. So then it hit me. I got back on the
ovirt engine I had installed from scratch, and it was fine installing
the node on cluster version 4.3 and the AMD EPYC IBPB SSBD. Then I just
had to follow your hint to import the Storage Domain (another issue
here, I tried to add it and then canceled, so it got stuck and couldn't
import it until I wen to psql and delete the line by hand).
The error message was always "Host ryzen-node moved to Non-Operational
state as host CPU type is not supported in this cluster compatibility
version or is not supported at all". And I never would guess that above
from this :(
When I tried to remove the node and add again, as it was CentOS8 based
and I could not install packages, it all went downhill to a state where
I tried to update packages from command line and the Engine GUI would
not work anymore.
Now the storm is behind me and I now have both node and engine up to
date. I will now seek information on how to keep them updated so I
don't have to be on that road again.
Thanks for all support!
I guess those rpm commands are no longer needed, right? If so, I can
send afterwards.
One last question, is it safe to update the cluster version to 4.6
using Ryzen 3400G?
thanks,
matheus
On Fri, Mar 25, 2022 at 20:31,
Nenhum_de_Nos<lojas(a)arroway.org>
wrote: On Fri, 25 Mar 2022 18:01:01 +0000 (UTC) Strahil Nikolov via
Users <users(a)ovirt.org> wrote:
> Even if you start with brand new engine you can still import the
> storage domains and the VMs inside it.Drawback -> you might have to
> delete any VM snapshots prior the import. Best Regards,Strahil
> Nikolov
Hi Strahil, thanks for the info.
My main issue now is the ovirt node. Its no easy for me to get an
intel based one and my ryzen based I cannot add to any cluster. I
tried creating clusters from 4.2 to 4.6, and its the same error over
and over again. Not compatible :(
I don't get it, before last satyrday it was, and worked fine. I
reinstalled all, I need those VM's ASAP, most my things are offline
for a week now. It's house stuff, yes, not big company production
VM's, no, but I need them to have all my stuff here.
thanks to all that helped.
matheus