[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Weird bridge problem Xen 4 + Debian Squeeze + HVM



Le vendredi 13 mai 2011 Ã 00:29 +0200, Francisco Javier Funes Nieto a
Ãcrit :
> Without the VM-HVM running...
> 
> 
> root@debian:~# brctl show
> bridge name     bridge id               STP enabled     interfaces
> eth0            8000.002215e17c2f       no              peth0
> 
> 
peth.... :/

Personnally, I stopped rely on the networks scripts of Xen  since uears.
As said in the Xen 4.1 release notes, thoose scripts are very fragiles
and I personnally find them evry complicated for what they really do:
configure the dom0 network. In the past I remember having some issue
related to arp being disabled on some underlying device or being nabled
whereas it should not... As you can see I don't clearly remember. :)
If you want to stick with the network-bridge scripts of xen ... may be
you can take a look at the ARP options of the devices. I would recommend
you let the OS take care of Dom0 network config (furthermore Xen 4.1
doesn't provides thoose scripts anymore i think).

I personnally use network-dummy as the network-script in xend-config.sxp
and configure Debian to setup a bridge at startup like this:

auto xenbr0
iface xenbr0 inet static
        address 192.168.x.y
        netmask 255.255.a.b
        broadcast 192.168.x.255
        gateway 192.168.x.z
        bridge_ports eth0
        bridge_fd 3
        bridge_maxwait 5

This config allow HVM (here a vista host) to access network with the
following in its configfile:

vif = [ 'bridge=xenbr0' ]

Hope it helps.

> 
> VM-HVM Running:
> 
> 
> root@debian:~# brctl show
> bridge name     bridge id               STP enabled     interfaces
> eth0            8000.002215e17c2f       no              peth0
>                                                         vif1.0
> 
> 
> If I use type=paravirtualized in the vif statement and add to the iso
> the GPLPV drivers (with nLite) the network works in the HVM domU
> (WinXP) but the default type=ioemu doesn't works.
> 
> 
> With Xen 3.4.2/Lenny dom0 it works flawlessly !! But Squeeze/Xen4
> doesn't.
> 
> 
> Any ideas?
> 
> 
> J.
> 2011/5/12 Alexandre Chapellon <a.chapellon@xxxxxxxxx>
>         you specify eth0 as the bridge interface?
>         does eth0 is really a bridge?
>         
>         what's the output of brctl show?
>         
>         Le jeudi 12 mai 2011 Ã 21:39 +0200, Francisco Javier Funes
>         Nieto a
>         Ãcrit :
>         
>         > Hi all,
>         >
>         >
>         > I've a Debian Squeeze 64bits box with Xen 4 from deb
>         repositories and
>         > PV/HVM support.
>         >
>         >
>         > root@debian:~# xm info |grep xen_caps
>         > xen_caps               : xen-3.0-x86_64 xen-3.0-x86_32p
>         hvm-3.0-x86_32
>         > hvm-3.0-x86_32p hvm-3.0-x86_64
>         >
>         >
>         > I'm trying to do a Windoze HVM but I've no network at all.
>         >
>         >
>         > This is the cfg file:
>         >
>         >
>         > #####################
>         > import os, re
>         > arch = os.uname()[4]
>         >
>         >
>         > kernel = "/usr/lib/xen-4.0/boot/hvmloader"
>         >
>         >
>         > builder='hvm'
>         > vcpus = '1'
>         > memory = 512
>         > shadow_memory = 8
>         > name = "xp1.canonigos.es"
>         > vif = [ 'type=ioemu, bridge=eth0' ]
>         > disk = [
>         >  'phy:/dev/mapper/vserver-xp1,ioemu:hda,w',
>         >  'file:/root/xp.iso,ioemu:hdb:cdrom,r'
>         > ]
>         >
>         >
>         >
>         >
>         > device_model = '/usr/lib/xen-4.0/bin/qemu-dm'
>         > # boot on floppy (a), hard disk (c) or CD-ROM (d)
>         > # default: hard disk, cd-rom, floppy
>         > boot="dc"
>         >
>         >
>         > sdl=0
>         > vnc=1
>         > vnclisten="0.0.0.0"
>         > vncconsole=1
>         > vncpasswd='canonigos'
>         > stdvga=0
>         > serial='pty'
>         > usbdevice='tablet'
>         >
>         >
>         > on_poweroff = 'destroy'
>         > on_reboot   = 'restart'
>         > on_crash    = 'restart'
>         >
>         >
>         > ########################
>         >
>         >
>         > With this config I've no network inside the domU-HVM but PV
>         domU's
>         > with Debian works great.
>         >
>         >
>         > If I change vif= statement to type=paravirtualized the
>         domU-HVM
>         > doesn't recognize the PCI Network device.
>         >
>         >
>         > There's some bug rolling' here?
>         >
>         >
>         > Thanks!
>         >
>         >
>         > J.
>         >
>         >
>         > --
>         > _____________________________________________
>         >
>         > Francisco Javier Funes Nieto [esencia@xxxxxxxxx]
>         > CANONIGOS
>         > Servicios InformÃticos para PYMES.
>         > Cl. Cruz 2, 1Â Oficina 7
>         > Tlf: 958.536759 / 661134556
>         > Fax: 958.521354
>         > GRANADA - 18002
>         >
>         
>         > _______________________________________________
>         > Xen-users mailing list
>         > Xen-users@xxxxxxxxxxxxxxxxxxx
>         > http://lists.xensource.com/xen-users
>         
>         --
>         horoa: la voie est libre
> 
> 
> 
> -- 
> _____________________________________________
> 
> Francisco Javier Funes Nieto [esencia@xxxxxxxxx]
> CANONIGOS
> Servicios InformÃticos para PYMES.
> Cl. Cruz 2, 1Â Oficina 7
> Tlf: 958.536759 / 661134556
> Fax: 958.521354
> GRANADA - 18002
> _______________________________________________
> Xen-users mailing list
> Xen-users@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-users

-- 
horoa: la voie est libre

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.