You didn't mention about the cpu_cap, if you have set any. That's the hard
limit.

I have encountered this issue before and in all cases, lowering the number
of vcpus, have resolved the issue.
Technically, the higher the vcpus in qemu, the more cycles it loses!

On Tue, Jan 31, 2017 at 5:14 PM, Jesus Cea <j...@jcea.es> wrote:

> This machine has 8 cores and I have configured vcpus=4 for a Solaris 10
> KVM zone.
> 
> When a non global native zone is "busy" (for instance, compiling a big
> project or doing video transcoding), the KVM Solaris 10 zone is unusable.
> 
> All zones have "cpu shares=100". I have raised the CPU shares for the
> KVM zone to 10000, it should get 99% of CPU when required, but I don't
> see any improvement. Same thing doing a "renice" to the QEMU process.
> 
> Nothing has worked so far.
> 
> I guess I could reserve 4 CPU cores for the KVM (cpu binding) but the
> actual CPU usage of the KVM is quite modest (like 2%) and I rather do
> not waste those 4 CPUs. Realistically I could set vcpu=1, but I don't
> think this is the real issue.
> 
> What I need is to guarantee that the KVM process gets the CPU when
> needed, with maximum priority, overriding other non global zones needs.
> 
> --
> Jesús Cea Avión                         _/_/      _/_/_/        _/_/_/
> j...@jcea.es - http://www.jcea.es/     _/_/    _/_/  _/_/    _/_/  _/_/
> Twitter: @jcea                        _/_/    _/_/          _/_/_/_/_/
> jabber / xmpp:j...@jabber.org  _/_/  _/_/    _/_/          _/_/  _/_/
> "Things are not so easy"      _/_/  _/_/    _/_/  _/_/    _/_/  _/_/
> "My name is Dump, Core Dump"   _/_/_/        _/_/_/      _/_/  _/_/
> "El amor es poner tu felicidad en la felicidad de otro" - Leibniz
> 



-------------------------------------------
smartos-discuss
Archives: https://www.listbox.com/member/archive/184463/=now
RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb
Powered by Listbox: http://www.listbox.com

Reply via email to