2010/10/20 Nikunj A. Dadhania <nikunj(a)linux.vnet.ibm.com>:
On Tue, 19 Oct 2010 15:49:05 +0200, Daniel Veillard
<veillard(a)redhat.com> wrote:
> On Tue, Oct 19, 2010 at 03:00:21PM +0200, Matthias Bolte wrote:
> > 2010/10/18 Nikunj A. Dadhania <nikunj(a)linux.vnet.ibm.com>:
[...]
> > > tools/virsh.c: Add new memory tunable "min_guarantee", currently
the user
> > > would be ESX.
> >
> > Well, who said ESX could support this? :)
> >
> > I didn't, I just said that you added min_guarantee to libvirt, but
> > didn't expose it in virsh.
>
> Hum ... I assumed there was an use case for it.
> On the other hand since it's now part of libvirt API as
> VIR_DOMAIN_MEMORY_MIN_GUARANTEE enum value, it's cleaner to have it
> available at the virsh level. If there is really no use (or no potential
> use in a reasonable future) for it, the best is to remove it altogether
> from the API and the of virsh, before next week release.
>
> Nikunj I could not find any reference to "the well-known tunable"
> min_guarantee (it's clearly not well known ...) but can you tell us
> where this comes from ?
>
VMWare:
=======
Reservation: Gauranteed lower bound on the amount of the physical memory that
the host reserves for the VM even in case of the overcommit. The
VM is allowed to allocate till this level and after it has hit
the reservation, those pages are not reclaimed. In case, if guest
is not using till the reservation, the host can use that portion
of memory.
And maybe in future cgroups.
Sorry, seems like I really did say that, but didn't remember doing so :(
Yes, reservation maps to min_guarantee :)
I'll add this to the ESX driver soon, so that we have at least one
user for each memtuneable.
Matthias