Re: [Xen-API] Comments on Xen API document

[Ewan Mellor <ewan@xxxxxxxxxxxxx>]

On Mon, Aug 21, 2006 at 05:27:04PM +0100, John Levon wrote:

> On Thu, Aug 17, 2006 at 12:13:24PM +0100, Ewan Mellor wrote:
> 
> > > VCPUs/params needs a detailed description of what it can/does contain
> > 
> > It's meant to be a free-form string, interpreted by whatever scheduler is
> > running.
> 
> So it's explicit that an XML RPC client can't configure the scheduler
> parameters, right?

That's not what I meant, no.  What I meant to say was that the client could
set scheduler parameters, but that it would need to understand whichever
scheduler is in use, and know how to encode the parameters for that.  In other
words, we need to state the meaning and format of the VCPUs/params field,
depending upon the scheduler in use.

In practice, for now at least, I think that just means defining the meaning of
this field for the credit scheduler, because that's will be the default for
Xen 3.0.3 and 3.0.4.   The intention for the longer term is to support the use
of other schedulers, but that's some way off yet.

> > > There's no mapping from VM vcpus to the host_cpu's it's pinned on.
> > 
> > Pinning is something that would get encoded into VCPUs/params, when the
> > scheduler supports such functionality.
> 
> More generally, though, there's no way to examine load balancing across the
> physical CPUs that I can see? This seems like pretty critical information.

That's a good point!  Thanks, I'll put that on the todo list.  I'll also put
together a fuller description of the CPU scheduler side of the API in general,
which certainly seems to be something that would be useful.

Cheers,

Ewan.

_______________________________________________
xen-api mailing list
xen-api@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-api