[pve-devel] Bug 1458 - PVE 5 live migration downtime degraded to several seconds (compared to PVE 4)

[Alexandre DERUMIER]

Just tested the patches on my old xeon, works fine, around 100ms of pause max.

But, it don't work will migration insecure (I think we should established the mtunnel also for insecure migration, to pass resume command)

----- Mail original -----
De: "aderumier" <aderumier at odiso.com>
À: "pve-devel" <pve-devel at pve.proxmox.com>
Envoyé: Mercredi 9 Août 2017 09:22:49
Objet: Re: [pve-devel] Bug 1458 - PVE 5 live migration downtime degraded to several seconds (compared to PVE 4)

>>I will see how we can speed upthe one with local storage a bit more as well, probably next week. 

I'm not sure how we can do faster pending write flush on block-job-complete. 

Maybe implementing COLO in the future (not sure it all parts are already implemented in qemu) could help to have both disk && memory migration synced. 


----- Mail original ----- 
De: "Fabian Grünbichler" <f.gruenbichler at proxmox.com> 
À: "pve-devel" <pve-devel at pve.proxmox.com> 
Envoyé: Mercredi 9 Août 2017 08:51:17 
Objet: Re: [pve-devel] Bug 1458 - PVE 5 live migration downtime degraded to several seconds (compared to PVE 4) 

On Tue, Aug 08, 2017 at 07:02:21PM +0200, Alexandre DERUMIER wrote: 
> Hi, 
> 
> I'm back from holiday, 
> 
> I'll test last mtunnel patches tomorrow . Thanks ! 

the relevant ones in qemu-server are already applied and packages are 
available in pve-no-subscription :) results in the forum indicate an 
improvement for shared storage migration, I will see how we can speed up 
the one with local storage a bit more as well, probably next week. 

_______________________________________________ 
pve-devel mailing list 
pve-devel at pve.proxmox.com 
https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel 

_______________________________________________ 
pve-devel mailing list 
pve-devel at pve.proxmox.com 
https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel