Hi Styve,

On Mon, Nov 23, 2015 at 3:10 PM, jaumotte, styve <s.jaumotte@maine-et-loire.fr> wrote:

Hi everybody,

 

After testing some features on 3.5, we are planning to finaly go to 3.6. Some problems still exist.

 

A major problem still remain on the « Highly Available » option on vm wich doesn’t work. I had a cluster with 4 engines and a simple vm. When I start a poweroff from the node where this vm is living, the node is shutting down but my vm doesn’t restart on another node of the cluster.


Do you have a fencing agent configured for your host? It is configureable under 'Edit Host'->'Power Management'. If there is no fencing agent configured, the engine can not make sure that the host is really off when it is non responsive. To avoid disk corruptions without a fencing agent, it does nothing. You can find more about HA here [1]. About fencing here [2].

The power managment of all the node are correctly configure. The HA feature of then hosted-engine is working well (except it is very long).


Glad to hear that this is working. Regarding the (sometimes long) downtimes, you can find some numbers here [3].

 

 

Another problem consist of passing usb host device to the virtual machine. We’ve got some specials usb keys for activating old application and we need to attach this key to vm. At first, I try with standard usb mass storage key to test this approach. I can’t start virtual machine when I add usb device, I always have the message « The host … did not satisfy internal filter HostDevice because it does not support host device passthrough ». Have any idea where I can find an HowTo to help me ?

 

Tanks for your help,

 

SJ

 


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


Best regards,
Roman

[1] http://www.ovirt.org/OVirt_Administration_Guide#.E2.81.A0Improving_Uptime_with_Virtual_Machine_High_Availability
[2] http://www.ovirt.org/OVirt_Administration_Guide#Host_Resilience
[3] http://www.ovirt.org/Hosted_Engine_Howto#What_is_the_expected_downtime_in_case_of_Datacenter_.2F_Host_.2F_VM_failure.3F