[Info-vax] VMware

[Grant Taylor]

On 12/10/19 8:14 AM, Craig A. Berry wrote:
> Doesn't being able to do that depend on OS capabilities that quiesce 
> everything while the running instance is being moved?

Nope.

The hypervisor does it transparently.

The guest VMs that have no idea that they are being moved any more than 
a physical box knows that it's being scooted across the floor while it's 
running with long network & power cords.

> I wouldn't expect a lot of the usual VM capabilities for OpenVMS 
> instances on first roll-out, but it would still be very nice to have 
> it running on VMWare, even if managing it is still mostly done the 
> old-fashioned way for now.

I'm of the opinion that there is very little perceptible difference 
(that matters) between physical hardware and virtualized machines.

I say "that matters" because there are usually some ways to tell if 
you're on a VM or a physical machine.  This is usually related to the 
textual descriptions of devices.  Does the OS /really/ care that the 
textual description is a Compaq SMAR Array / DGD (whatever the old DEC / 
Compaq SAN was) vs a VMware virtual disk?

So, if it works on physical hardware, I'd expect it has a very good 
chance (> 80%) that it will work the same way on virtual hardware.

The biggest difference that I see (other than textual descriptions) are 
the emulated devices.  Different hypervisors emulate different SCSI 
controllers or video cards.  But if you have drivers for the respective 
(physical / virtual) hardware, it shouldn't matter.



-- 
Grant. . . .
unix || die