On Mon, Jan 25, 2016 at 2:18 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
On Wed, Jan 20, 2016 at 12:59 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
The oVirt Project is pleased to announce the availability
of the Third Release Candidate of oVirt 3.6.2 for testing, as of January 20th, 2016



Tested successfully on CentOS 7.2 with SH Engine on NFS and verified correct auto-import of SH Domain.
Hypervisor is a CentOS 7.2 VM inside Qemu/KVM in virt-manager (under my laptop that is Fedora 23).
So it is a nested virtualization environment.

Some NOTES
1) appliance
I see that the appliance rpm pulled in was
 ovirt-engine-appliance-3.6-20151216.1.el7.centos.noarch

I presumed it was not populated with correct rpms.
So in "hosted-engine --deploy" setup I used the appliance, but answered

Automatically execute engine-setup on the engine appliance on first boot (Yes, No)[Yes]? No

I then connect to the appliance OS and configure 3.6-pre repo and run 

yum update

Strange that this pulled in the package

ovirt-release36-003-1.noarch

and so the repo file was overwritten and was again 3.6 only and not 3.6-pre.
I re-modified /etc/yum.repos.d/ovirt-3.6.repo re-adding the pre.

and finally
engine-setup

and successfully completed the final part on host.

2) localtime on Engine was not asked but was set to UTC.
So
[root@shengine ~]# ln -sf /usr/share/zoneinfo/Europe/Rome /etc/localtime 

shutdown engine OS and after some minutes the VM came back online.


You are right here: we are not copying the host timezone to the appliance: we can do it via cloud-init.
Good catch!
 
3) As this is a nested environment with NFS domain provided by the only host itself,
In /usr/lib/systemd/system/ovirt-ha-broker.service

Agdded in section [Unit]

After=nfs-server.service
So that I can manage shutdown/restart of host for complete maintenance operations

4) After adding first data domain, the hosted_storage storage domain came up automatically without any problem. Able to see engine vm in web admin portal and its disk and access its console (used vnc)
Also able to add ISO Domain.
(both data and iso are NFS on the host itself)
 
5) The only "problem" (?) I see is this message that I don't understand.
On hypervisors

[root@ovc72 ~]# systemctl status ovirt-ha-agent -l
● ovirt-ha-agent.service - oVirt Hosted Engine High Availability Monitoring Agent
   Loaded: loaded (/usr/lib/systemd/system/ovirt-ha-agent.service; enabled; vendor preset: disabled)
   Active: active (running) since Mon 2016-01-25 11:02:56 CET; 1h 0min ago
 Main PID: 17138 (ovirt-ha-agent)
   CGroup: /system.slice/ovirt-ha-agent.service
           └─17138 /usr/bin/python /usr/share/ovirt-hosted-engine-ha/ovirt-ha-agent --no-daemon

Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.lib.storage_server.StorageServer:Connecting storage server
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.lib.storage_server.StorageServer:Refreshing the storage domain
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine:Preparing images
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.lib.image.Image:Preparing images
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine:Reloading vm.conf from the shared storage domain
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: INFO:ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine.config:Trying to get a fresher copy of vm configuration from the OVF_STORE
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: WARNING:ovirt_hosted_engine_ha.lib.ovf.ovf_store.OVFStore:Unable to find OVF_STORE
Jan 25 12:03:27 ovc72.localdomain.local ovirt-ha-agent[17138]: ovirt-ha-agent ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine.config ERROR Unable to get vm.conf from OVF_STORE, falling back to initial vm.conf


We will let you edit some of the engine VM parameters from the engine itself.
All the modification will be saved as OVF file on a special volume called OVF_STORE; when needed ovirt-ha-agent tries to get the up-to-date engine VM configuration from there and if there is any issue it falls back to the initial vm.conf from the setup time.
The issue is that by default OVF_STORE gets populated only after 1 hour and so if you wait less than 1 hour you'll get that message since the OVF_STORE is still not there.
 
6) ISO upload from engine by default doesn't work because its own host doesn't allow it
[root@shengine ~]# ovirt-iso-uploader -i ISO_DOMAIN upload /root/CentOS-7-x86_64-NetInstall-1511.iso 
Please provide the REST API password for the admin@internal oVirt Engine user (CTRL+D to abort): 
Uploading, please wait...
ERROR: mount.nfs: Connection timed out

If on host I run (.73 is IP of engine)
# iptables -I INPUT -s 192.168.122.73 -j ACCEPT

Then on engine:
[root@shengine ~]# ovirt-iso-uploader -i ISO_DOMAIN upload /root/CentOS-7-x86_64-NetInstall-1511.iso 
Please provide the REST API password for the admin@internal oVirt Engine user (CTRL+D to abort): 
Uploading, please wait...
INFO: Start uploading /root/CentOS-7-x86_64-NetInstall-1511.iso 
Uploading: [########################################] 100%
INFO: /root/CentOS-7-x86_64-NetInstall-1511.iso uploaded successfully

7) Installed a CentOS 7.2 VM with netinst.iso. Access to spice console ok.
(85MB/s when downloading on nested network not so bad... ;-)

Very good starting point!

Gianluca


_______________________________________________
Announce mailing list
Announce@ovirt.org
http://lists.ovirt.org/mailman/listinfo/announce