also here:
https://patchwork.kernel.org/patch/5825741/


Seem that increasing regions decrease performance, because of extra lookups.
(I think with more regions, it's some kind of tree, where your need to lookup 
first the parent region, then the child region)


----- Mail original -----
De: "aderumier" <aderum...@odiso.com>
À: "pve-devel" <pve-devel@pve.proxmox.com>
Cc: "dietmar" <diet...@proxmox.com>
Envoyé: Lundi 12 Décembre 2016 19:42:03
Objet: Re: [pve-devel] qemu memory hotplug limited to 64 dimm

I found some kernel patch to bump it to 509 (seem that another kvm fixed was 
needed) 

http://www.spinics.net/lists/kvm/msg117654.html 

But I think it was never commited to the kernel. 

I'll try follow the discuss. 

----- Mail original ----- 
De: "aderumier" <aderum...@odiso.com> 
À: "dietmar" <diet...@proxmox.com> 
Cc: "pve-devel" <pve-devel@pve.proxmox.com> 
Envoyé: Lundi 12 Décembre 2016 19:29:54 
Objet: Re: [pve-devel] qemu memory hotplug limited to 64 dimm 

>>Can't we simply increase that max_mem_regions default value? 

I'm trying to see why they have limited it to 64. 

maybe a qemu bug with previous qemu version? don't known. 

I have some big spare servers to test with 370G ram, i'll try to do test with 
them. 


----- Mail original ----- 
De: "dietmar" <diet...@proxmox.com> 
À: "aderumier" <aderum...@odiso.com>, "pve-devel" <pve-devel@pve.proxmox.com> 
Envoyé: Lundi 12 Décembre 2016 17:09:41 
Objet: Re: [pve-devel] qemu memory hotplug limited to 64 dimm 

Can't we simply increase that max_mem_regions default value? 


> On December 12, 2016 at 4:54 PM Alexandre DERUMIER <aderum...@odiso.com> 
> wrote: 
> 
> 
> vhost 64 limit has been introduced here, because of qemu crash with bigger 
> value 
> 
> https://patchwork.kernel.org/patch/6709021/ 
> 
> ----- Mail original ----- 
> De: "aderumier" <aderum...@odiso.com> 
> À: "pve-devel" <pve-devel@pve.proxmox.com> 
> Envoyé: Lundi 12 Décembre 2016 16:51:11 
> Objet: [pve-devel] qemu memory hotplug limited to 64 dimm 
> 
> Hi, 
> 
> a proxmox user has reported problem to hotplug memory with high values (48G) 
> 
> https://forum.proxmox.com/threads/apt-get-upgrade-to-update-proxmox-server.30992/
>  
> 
> 
> 
> Seem that qemu check about a vhost value 
> 
> https://lists.nongnu.org/archive/html/qemu-devel/2015-07/msg05998.html 
> 
> cat /sys/module/vhost/parameters/max_mem_regions > 64 (verified on my 
> servers). 
> 
> 
> I think this limit was not present when we have implemented hotplug. 
> 
> Not sure was we can do, but I think our method don't scale anymore :( 
> 
> Maybe should we introduce a new fixed "dimm size" option, to allow hotplug of 
> more memory ? 
> 
> 
> 
> _______________________________________________ 
> pve-devel mailing list 
> pve-devel@pve.proxmox.com 
> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel 
> 
> _______________________________________________ 
> pve-devel mailing list 
> pve-devel@pve.proxmox.com 
> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel 

_______________________________________________ 
pve-devel mailing list 
pve-devel@pve.proxmox.com 
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel 

_______________________________________________
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel

Reply via email to