Re: [Users] vms behaviour when a host shut down

2014-01-26 Thread Itamar Heim

On 01/24/2014 04:43 PM, Nathanaël Blanchet wrote:

Typically, we ask to our inverter to send a signal to the host to shut
down when it runs on his battery so as OS extinction to be clean. So
we also want that the guests to shut down cleanly because they are so
much important than the host, to prevent of database corruption for
instance. For this, we could imagine host sending an init signal through
a virtual serial port to the VMs. Do you think it is possible in that or
an other way?


you should send the engine an API call to either shutdown the VMs, or 
migrate them to another host.

moving the host to maint before killing it is also a nicer behavior.
if the VMs will die, no via the engine, then the HA ones will be 
restarted on another host.

live migrating them if possible would be better than kill/restart.
(but if you ask engine to stop them, rather than migrate them, then you 
also have to own starting them again)



Le 24/01/2014 15:17, Dan Kenigsberg a écrit :

On Fri, Jan 24, 2014 at 10:01:52AM +0100, Nathanaël Blanchet wrote:

Hi all,

I'd like to be sure about the behaviour of vms when hist host goes
down for any reason. I suppose high availibility is for migrating
vms to another host, but how to do if I prefer them to shut down at
the same time of the host?

For a planned shutdown of a host, you should move it to maintenance
mode (which migrates all VMs from it) or shut down the VMs individually.

If the host unexpectedly dies, it would be fenced by Engine.
Highly-available VMs that were running on it, would be re-started on
another host. You can mimic such a condition with logging into the host
as root and doing
 service vdsmd stop
 pkill qemu-kvm
(and may the force be with your guest's data)


--
Nathanaël Blanchet

Supervision réseau
Pôle exploitation et maintenance
Département des systèmes d'information
227 avenue Professeur-Jean-Louis-Viala
34193 MONTPELLIER CEDEX 5   
Tél. 33 (0)4 67 54 84 55
Fax  33 (0)4 67 54 84 14
blanc...@abes.fr



___
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


[Users] vms behaviour when a host shut down

2014-01-24 Thread Nathanaël Blanchet

Hi all,

I'd like to be sure about the behaviour of vms when hist host goes down 
for any reason. I suppose high availibility is for migrating vms to 
another host, but how to do if I prefer them to shut down at the same 
time of the host?


--
Nathanaël Blanchet

Supervision réseau
Pôle exploitation et maintenance
Département des systèmes d'information
227 avenue Professeur-Jean-Louis-Viala
34193 MONTPELLIER CEDEX 5   
Tél. 33 (0)4 67 54 84 55
Fax  33 (0)4 67 54 84 14
blanc...@abes.fr

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


Re: [Users] vms behaviour when a host shut down

2014-01-24 Thread Meital Bourvine
Take a look at the configuration options: when editing a vm, press on Choose 
Advanced Optionsbutton - see Host and High Availability tabs. 

- Original Message -

 From: Nathanaël Blanchet blanc...@abes.fr
 To: users users@ovirt.org
 Sent: Friday, January 24, 2014 9:01:52 AM
 Subject: [Users] vms behaviour when a host shut down

 Hi all,

 I'd like to be sure about the behaviour of vms when hist host goes down for
 any reason. I suppose high availibility is for migrating vms to another
 host, but how to do if I prefer them to shut down at the same time of the
 host?
 --
 Nathanaël Blanchet

 Supervision réseau
 Pôle exploitation et maintenance
 Département des systèmes d'information
 227 avenue Professeur-Jean-Louis-Viala
 34193 MONTPELLIER CEDEX 5
 Tél. 33 (0)4 67 54 84 55
 Fax  33 (0)4 67 54 84 14 blanc...@abes.fr

 ___
 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


Re: [Users] vms behaviour when a host shut down

2014-01-24 Thread Dan Kenigsberg
On Fri, Jan 24, 2014 at 10:01:52AM +0100, Nathanaël Blanchet wrote:
 Hi all,
 
 I'd like to be sure about the behaviour of vms when hist host goes
 down for any reason. I suppose high availibility is for migrating
 vms to another host, but how to do if I prefer them to shut down at
 the same time of the host?

For a planned shutdown of a host, you should move it to maintenance
mode (which migrates all VMs from it) or shut down the VMs individually.

If the host unexpectedly dies, it would be fenced by Engine.
Highly-available VMs that were running on it, would be re-started on
another host. You can mimic such a condition with logging into the host
as root and doing
service vdsmd stop
pkill qemu-kvm
(and may the force be with your guest's data)
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] vms behaviour when a host shut down

2014-01-24 Thread Nathanaël Blanchet
Typically, we ask to our inverter to send a signal to the host to shut 
down when it runs on his battery so as OS extinction to be clean. So 
we also want that the guests to shut down cleanly because they are so 
much important than the host, to prevent of database corruption for 
instance. For this, we could imagine host sending an init signal through 
a virtual serial port to the VMs. Do you think it is possible in that or 
an other way?

Le 24/01/2014 15:17, Dan Kenigsberg a écrit :

On Fri, Jan 24, 2014 at 10:01:52AM +0100, Nathanaël Blanchet wrote:

Hi all,

I'd like to be sure about the behaviour of vms when hist host goes
down for any reason. I suppose high availibility is for migrating
vms to another host, but how to do if I prefer them to shut down at
the same time of the host?

For a planned shutdown of a host, you should move it to maintenance
mode (which migrates all VMs from it) or shut down the VMs individually.

If the host unexpectedly dies, it would be fenced by Engine.
Highly-available VMs that were running on it, would be re-started on
another host. You can mimic such a condition with logging into the host
as root and doing
 service vdsmd stop
 pkill qemu-kvm
(and may the force be with your guest's data)


--
Nathanaël Blanchet

Supervision réseau
Pôle exploitation et maintenance
Département des systèmes d'information
227 avenue Professeur-Jean-Louis-Viala
34193 MONTPELLIER CEDEX 5   
Tél. 33 (0)4 67 54 84 55
Fax  33 (0)4 67 54 84 14
blanc...@abes.fr

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