[pve-devel] qemu-server : cpu hotplug rework

Alexandre DERUMIER aderumier at odiso.com
Fri Jan 23 07:42:30 CET 2015


>>Or do we need to start the VM with exactly the same topology on the other side?

Yes, you need same sockets/cores/(threads) and numa on each side, because guest is aware of that,
so it'll crash after migration if different.

----- Mail original -----
De: "dietmar" <dietmar at proxmox.com>
À: "aderumier" <aderumier at odiso.com>
Cc: "pve-devel" <pve-devel at pve.proxmox.com>
Envoyé: Vendredi 23 Janvier 2015 06:54:59
Objet: Re: [pve-devel] qemu-server : cpu hotplug rework

> >>Or should we simply assume 'vcpus' is the switch to enable hotplug? 
> Yes, I thinked at something like this. 
> 
> On gui, on processor form, 
> 
> maybe can we replace the "total cores: x" text, by a numberfield (vcpus) 
> 
> "total enabled cores: x" 
> 
> -default is (sockets * cores) 

Do you know if CPU topology information is included in live migration data 
stream? 
Or do we need to start the VM with exactly the same topology on the other side? 


More information about the pve-devel mailing list