Right fresh install of CentOS 7.4 all updated

Installed ovirt using the repositories from provided by ovirt-release42.rpm

The install seems to go fine, once the hosted engine is up /var/log/message is getting filled with the following messages:

journal: ovirt-ha-agent ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine.config ERROR Unable to identify the OVF_STORE volume, falling back to initial vm.conf. Please ensure you already added your first data domain for regular VMs


I've added a new storage domain as used to be required, but as Nir suggested earlier the expected behaviour in 4.2 is that the hosted_storage domain remains master and /var/log/messages continues to fill with the above messages.



One other niggle I've found in the current installer is the space required to generate hosted_engine, the instructions on this page Deploying Self-Hosted Engine suggest that the installer will check that there is enough available space or prompt for an alternative location if there isn't enough at /var/tmp.
Actually it does neither and fails the deployment only saying to check the log files for a possible cause.



On 18 March 2018 at 11:12, Maton, Brett <matonb@ltresources.co.uk> wrote:
Ok, that does make sense.

I'm rebuilding again at the moment anyway, I'll keep this thread updated as deployment progresses.

On 18 March 2018 at 10:55, Nir Soffer <nsoffer@redhat.com> wrote:
On Sun, Mar 18, 2018 at 12:01 PM Maton, Brett <matonb@ltresources.co.uk> wrote:
Yes, both are on NFS and both are mounted.

For what it's worth all the storage this cluster will connect to is on the same NAS device, same permissions etc etc, and it does connect it's just not 'flipping' the master domain.

This is the first time I've tried a fresh install of 4.2.1.7-1.el7.centos, I generally in-place upgrade on this cluster.

 If there are any logs they might give a clue as to what's happening I'm happy to shre those.

In the mean time I'll flatten the host and deploy from scratch again.

No need to flatten the host, I think the behavior is expected in 4.2. The installation process
was streamlines and there is no need now to create another storage domain to force the
system to import the hosted engine domain.

Since the hosted engine storage domain cannot be deactivated, using it for master is good.
it meas you can deactivate any other storage domain if needed.

Maybe the documentation needs update?

On 17 March 2018 at 18:54, Juan Pablo <pablo.localhost@gmail.com> wrote:
as soon as you install ovirt when you add a second storage domain, it should switch to it as the master, not hosted storage as you have. are both nfs?

JP


2018-03-17 15:45 GMT-03:00 Maton, Brett <matonb@ltresources.co.uk>:
I have added a new domain, but it's not playing ball.

hosted_storage Data (master)  Active
vm_storage     Data           Active

On 17 March 2018 at 17:24, Juan Pablo <pablo.localhost@gmail.com> wrote:
you need to configure first a domain (can be temporary) then, import the storage domain.



2018-03-17 7:13 GMT-03:00 Maton, Brett <matonb@ltresources.co.uk>:
I've just reinstalled ovirt 4.2 (ovirt-release42.rpm) on a clean CentOS 7.4 host, all went fine however

Now that the hosted engine is up, I've added a data storage domain but it looks like the next step that detects / promotes the new data domain to master isn't being triggered.

Hosted engine and data domain are on NFS storage, /var/log/messages is filling with these messages

journal: ovirt-ha-agent ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine.config ERROR Unable to identify the OVF_STORE volume, falling back to initial vm.conf. Please ensure you already added your first data domain for regular VMs


I ran into the same issue when I imported an existing domain.

_______________________________________________
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users





_______________________________________________
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users