> Hi,
>
> I'm trying to run Windows XP guest on a b90 host. No
> matter what I've tried doing, I always get "Device 0
> (vif) could not be connected. Backend device not
> found." when I try creating a guest with a vif.
>
> Unfortunately, the host uses an Asus P5K-VM
> motherboard with Marvell 88E8056 PCI-E Gigabit
> Ethernet Controller, and had to go through some bells
> and hoops to make it work (see link to bells and
> hoops article below), including the installation of a
> driver from SysKonnect's website.
>
> I feel a bit sheepish for not being able to find the
> proper explanation for Xen and networking in Solaris
> (is there such a thing?). But from what I collected
> on the Internet I /think/ the driver may be an old
> GLDv2, and Xen requires GLDv3 architected drivers to
> work with network. Is that correct?
Yes
>
> Finally, I wanted to give up on 'real' networking
> from my machine and let Solaris route the traffic
> using something akin to 'host only networking', and
> still couldn't find the proper explanation for Xen
> and networking in Solaris. What other vif types are
> there other than ioemu and bridge, and where do I
> read about them?
>
> Bottom line is: what's the path you'd recommend to
> get me a running Windows XP box with networking?
Buy RTL 8169SC (1) or RTL 8139 (2) and plug card into PCI slot
Reinstall Nevada.
I guess :-
(1)
bash-3.2# dladm show-link
LINK CLASS MTU STATE OVER
rge0 (?) phys 1500 up --
(2)
bash-3.2# dladm show-link
LINK CLASS MTU STATE OVER
rlts0 phys 1500 up --
svccfg -s xvm/xend setprop config/vncpasswd = astring: \"secret\"
svccfg -s xvm/xend setprop config/vnc-listen = astring: \"0.0.0.0\"
svccfg -s xvm/xend setprop config/default-nic="rge0" (" rlts0")
svcadm refresh xvm/xend
svcadm restart xvm/xend
Virt-install WinXP.
Realtek adapters work for sure.
>
> Thanks,
> - Yaniv
>
> (bells and hoops galore
> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bu
> g_id=6546335)
This message posted from opensolaris.org
_______________________________________________
xen-discuss mailing list
[email protected]
