Thank you for clarification!
Am 02.11.2017 um 20:55 schrieb Arman Khalatyan:
Ovirt HA means that if you have virtual machine running on the
ovirts
environment(let us say 2 nodes) then if the bare metal gets troubles,
VM will be restarted on the second one, the failed host must be
fenced: poweroff/reboot, but the HA model assumes that the both bare
metal machines are always on and healthy. if second host is off, then
simply you dont have HA, you should ask someone to turn on the second
host in order to rerun your VMs.:)
Usually if you turn off the "healthy host" it does not have any
information to sync, the ovirt-engine manages all things.
Ok, then HA is not the
right choice.
(Maybe question belongs to the wrong forum?)
The ovirt does not contain any sync / HA functionality in the data side.
Maybe you are looking for some ha/failover-file systems like a
glusterfs(geo-replication) or drbd(real-time replication) or
zfs: send receive(smart backups+snapshots) or some thing similar.
When I understand
your right, then there is no necessary data on the
nodes, all information have the ovirt engine? My VM images are on a nfs
share, at the moment.
When one node crashes I can just migrate the VM to the second node? That
would be wonderful!
On Thu, Nov 2, 2017 at 8:20 PM, Jonathan Baecker
<jonbae77(a)gmail.com> wrote:
> Hello everybody,
>
> I would like to sync two nodes, but I want that only one node runs
> permanently. Only once a week or a month I want to start the second node and
> sync them again, if this is necessary.
>
> What you would recommend for this scenario? oVirt have a HA functions, what
> I could use, but I thought oVirt brings then maybe errors when one node is
> always off. I'm wrong here? Or is there other options what works better?
>
> Have a nice day!
>
> Jonathan
>
> _______________________________________________
> Users mailing list
> Users(a)ovirt.org
>
http://lists.ovirt.org/mailman/listinfo/users