Re: NetBSD domU and the real server

["Dirk H. Schulz" <dirk.schulz%kinzesberg.de@localhost>]

Hi Jean-Yves,

Jean-Yves Migeon schrieb:
> I am assuming you are not using PCI passthru, or HVM.
No, neither of those.
> domU is completely hardware independent, except for the machine arch 
> (x86). It is expected to be completely isolated (by design) from the 
> hardware, and cannot access devices, chipsets, or memory. Literally 
> everything has to go and get validated through dom0 and/or Xen.
>
> dom0 is critical in regards to security, and must remain 100% reliable 
> even in cases where domU is crippled (security by isolation). If it 
> crashes because of a domU, it is clearly buggy, and makes assumption 
> on a state or condition it should not to.
That is the theory I also had in mind. :-)
> Faulty hardware may be a source of problem too.
The real server ran CentOS dom0 and domU without any crashes for many, 
many months now. I simply reinstalled it with Debian Lenny and started 
moving Debian and NetBSD domUs there.
That means: It can be a hardware failure, of course, but it is unlikely, 
I think, especially with this behaviour pattern: a crash only if certain 
domUs are run.
> Without further information, it is hard to tell. A good start would be 
> a stacktrace, xm dmesg, or Xen console output.
A stacktrace of the dom0 kernel, then?
I have never done that (and googling did not show up any fast howto), so 
please point me to docs on that, if possible.


Dirk