On Mon, 24 Oct 2005, Nivedita Singhvi wrote:
> Anthony.Golia@xxxxxxxxxxxxxxxxx wrote:
>
> > thx. that helps but leaves me with a few more questions. any doc is
> > appreciated. i had bridging on then changed the xend-config.sxp
> > to use vif-route so now i have a completely different set of ints
> > (see below). i thought the eth0 in domU was linked to vifx.U in dom0?
> > so with routing on there's no peth0? is eth0 and eth1 still my real,
> > physical interface? the ascii diagram you have above is helpfull how does
> > it look with routing?
> >
> > bash-3.00# ip addr
> > 1: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
> > link/ether 00:09:6b:b5:6b:94 brd ff:ff:ff:ff:ff:ff
> > inet 172.31.205.177/24 brd 172.31.205.255 scope global eth0
> > 2: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
> > link/ether 00:09:6b:b5:6b:95 brd ff:ff:ff:ff:ff:ff
> > inet 172.31.215.177/24 brd 172.31.215.255 scope global eth1
> > 3: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
> > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
> > inet 127.0.0.1/8 scope host lo
> > inet 172.31.254.207/32 brd 172.31.254.207 scope global lo:hostname
> > 4: vif0.0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop
> > link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
> > 5: veth0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop
> > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
> > 6: vif1.1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue
> > link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
> > inet 169.254.1.0/32 brd 169.254.1.0 scope global vif1.1
>
> This really depends on your patchset/version you are running, and
> your local script. Are you running the default network-route and
> vif-route that's currently in the unstable tree? If you change
> the vif file to be vif-route, you should have changed the network
> file to be network-route, too. Looks like you have two interfaces
> in dom0, that's why you have both eth0 and eth1. You have two
> virtual backends (vif0.0, vif1.1). The default route scripts don't
> rename the interface (so you won't have a peth0, for instance)..
> Is the above working for you? Can you send netstat -rn output?
>
> In brief:
>
> physical interface (e.g. eth0) <=> virtual backend (e.g. vif0.0)
> ||
> \\
> =====> domU nic (eth0)
>
thx. running changeset: 7421:aabc33c3c0ac with the default vif-route and
network-route. i dont fully understand the
tie between vif0.0 and eth0 in dom0. i thought vif1.1 in dom0 was linked
to eth0 in domU (not vif0.0 in dom 0 as in ur diagram). yes i have two
real ints eth0 and eth1 (and i intend to use both of them using linux's
advanced routing/load balancing).
bash-3.00# ip route show
172.31.254.253 dev vif1.1 scope link < ***** i added this manually
172.31.215.128/25 dev eth1 proto kernel scope link src 172.31.215.177
172.31.205.128/25 dev eth0 proto kernel scope link src 172.31.205.177
172.31.205.0/24 dev eth0 proto kernel scope link src 172.31.205.177
172.31.215.0/24 dev eth1 proto kernel scope link src 172.31.215.177
169.254.0.0/16 dev eth1 scope link
default
nexthop via 172.31.205.1 dev eth0 weight 1
nexthop via 172.31.215.1 dev eth1 weight 1
despite my confusion (getting better thx to you) things seem to be
working, mostly, cept some small thingies like:
- xm create doesnt seem to like a 32 bit netmask (255.255.255.255)
in the dom conf file
- same for having dom1 be its own default router (i.e. proxy arp)
> I'll confirm this once I get the latest tip working...
>
> thanks,
> Nivedita
>
Cheers,
Anthony
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home