[prev in list] [next in list] [prev in thread] [next in thread] 

List:       opensolaris-discuss
Subject:    Re: [osol-discuss] Virtualization on solaris/opensolaris
From:       Florian Ermisch <floh.edo () googlemail ! com>
Date:       2010-07-28 11:51:53
Message-ID: 4C5019D9.3080600 () googlemail ! com
[Download RAW message or body]

Am 28.07.10 11:40, schrieb Matthias Pfützner:
[...]
So, in order to minimize that problem, "thin" hypervisors have been 
designed,
> like Xen. Idea is: Less code, less faults. Compare VMware vs. Xen, and you'll
> understand, what I'm talking about... In VMware, there's way more code (as it
> is a thick hypervisor), so the probability of an error happeing inside thatr
> code is higher then with Xen... Then we have PVM and HVM modes. In PVM mode,
> there's even less code from the hyperviso in use, as the guest OS can access
> the udnerlying hardware in parts directly, whereas in HVM mode, all access to
> the underlying hardware is abstracted by the hypervisor. Again: Higer
> probability for errors in software-layers happening...
>
>       Matthias

Actually the abstracion of "hardware" for paravirtualized domains is 
higher. In HVM mode the guest OS still works on something which looks 
and tries to behave like hardware, i.e. an Intel e1000g0 NIC. In PVM 
mode the guest OS only uses interfaces of the hypervisor which provide 
the functionality of a class of hardware.
but still less code: a reduced set of drivers in the guest OS, which is 
or at least can be tailored for the hypervisor.

scnr ;)

Regards, Florian
_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org

[prev in list] [next in list] [prev in thread] [next in thread]