veth0 problems and network-route

veth0 problems and network-route

[wei huang]

Hi,

After working on something else for a while, it is the time that I have to
solve the network problem wrt dom0 and domU.

I posted some emails about this problem before but it got overwhelmed. So
I re-post it here:
I am using AS4 linux with broadcom tg3 gigabit ethernet.

1) I am using the newest code in xen-unstable. Everything works smoothly
until I call 'xend start'. After that the network traffic just dies. I've
looked at the ifconfig before and after xend start, every seems normal:
veth0 gets the ip and mac address from eth0. And eth0's Mac is set to
fe:ff:ff:ff:ff:ff. xen-br0 is created and vif0.0 and veth0 is attached to
it. Actually I should say that it is exactly the same sympton as bug 62 in
your bugzilla. Is that problem solved? Am I miss something?


2) I stuck it there for some while. Then I find there is another path in
the network script:
else
        transfer_addrs ${netdev} ${bridge}
        transfer_routes ${netdev} ${bridge}
fi
I believe this is the old way to setup network in xen-2.0 (are they?).
Anyway, I try to go through that path and things seem going fine. I can
start domU and domU even gets its ip address from dhcp. Both domU and dom0
can connect outside. But domU and dom0 just cannot talk to each other!
They can ping each other but ssh and socket doesn't work...
Following is ifconfig of domU (to save space, I only copy the address
lines):
eth0      Link encap:Ethernet  HWaddr AA:00:00:00:00:11
          inet addr:9.2.72.107  Bcast:9.2.72.127  Mask:255.255.255.128
lo        Link encap:Local Loopback
          inet addr:1 overr27.0.0.1  Mask:255.0.0.0
And the dom0:
eth0      Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
          inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
vif1.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
xen-br0   Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
          inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
And the output of 'netstat -nr' in dom0:
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt  Iface
9.2.72.0        0.0.0.0         255.255.255.128 U         0 0          0 xen-br0
169.254.0.0     0.0.0.0         255.255.0.0     U         0 0          0 eth0
0.0.0.0         9.2.72.1        0.0.0.0         UG        0 0          0 xen-br0

Do you have any idea what is happening? And if you need more information
please suggest me.

Thank you very much!

Regards,
Wei Huang

774 Dreese Lab, 2015 Neil Ave,
Dept. of Computer Science and Engineering
Ohio State University
OH 43210
Tel: (614)292-8501

Re: veth0 problems and network-route

[Florian Gödden]

wei huang wrote:

>xen-br0 is created and vif0.0 and veth0 is attached to
>it. 
>
Is veth0 really attached to the bridge? If i am right that will produce 
a loop. Or do you mean eth0 instead of veth0?

MfG

Florian

Re: veth0 problems and network-route

[wei huang]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: TEXT/PLAIN; charset=X-UNKNOWN, Size: 656 bytes --]

Sorry, you are right, I mean eth0

Regards,
Wei Huang

774 Dreese Lab, 2015 Neil Ave,
Dept. of Computer Science and Engineering
Ohio State University
OH 43210
Tel: (614)292-8501


On Fri, 29 Jul 2005, [ISO-8859-1] Florian Gödden wrote:

> wei huang wrote:
>
> >xen-br0 is created and vif0.0 and veth0 is attached to
> >it.
> >
> Is veth0 really attached to the bridge? If i am right that will produce
> a loop. Or do you mean eth0 instead of veth0?
>
> MfG
>
> Florian
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>

Re: veth0 problems and network-route

[wei huang]

Hi,

I tried xen-2.0 and found that the network works pretty smoothly. It seems
that the veth0 stuff is put into the unstable version recently.

Do you have any back port patch to the version vnet which doesn't use
veth0? It might help in my case. PS: I need to run program with some new
features in unstable, so I have to stick to the unstable versoin.

Thank you very much!

Regards,
Wei Huang


On Thu, 28 Jul 2005, wei huang wrote:

> Hi,
>
> After working on something else for a while, it is the time that I have to
> solve the network problem wrt dom0 and domU.
>
> I posted some emails about this problem before but it got overwhelmed. So
> I re-post it here:
> I am using AS4 linux with broadcom tg3 gigabit ethernet.
>
> 1) I am using the newest code in xen-unstable. Everything works smoothly
> until I call 'xend start'. After that the network traffic just dies. I've
> looked at the ifconfig before and after xend start, every seems normal:
> veth0 gets the ip and mac address from eth0. And eth0's Mac is set to
> fe:ff:ff:ff:ff:ff. xen-br0 is created and vif0.0 and veth0 is attached to
> it. Actually I should say that it is exactly the same sympton as bug 62 in
> your bugzilla. Is that problem solved? Am I miss something?
>
>
> 2) I stuck it there for some while. Then I find there is another path in
> the network script:
> else
>         transfer_addrs ${netdev} ${bridge}
>         transfer_routes ${netdev} ${bridge}
> fi
> I believe this is the old way to setup network in xen-2.0 (are they?).
> Anyway, I try to go through that path and things seem going fine. I can
> start domU and domU even gets its ip address from dhcp. Both domU and dom0
> can connect outside. But domU and dom0 just cannot talk to each other!
> They can ping each other but ssh and socket doesn't work...
> Following is ifconfig of domU (to save space, I only copy the address
> lines):
> eth0      Link encap:Ethernet  HWaddr AA:00:00:00:00:11
>           inet addr:9.2.72.107  Bcast:9.2.72.127  Mask:255.255.255.128
> lo        Link encap:Local Loopback
>           inet addr:1 overr27.0.0.1  Mask:255.0.0.0
> And the dom0:
> eth0      Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
>           inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
> lo        Link encap:Local Loopback
>           inet addr:127.0.0.1  Mask:255.0.0.0
> vif1.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
> xen-br0   Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
>           inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
> And the output of 'netstat -nr' in dom0:
> Kernel IP routing table
> Destination     Gateway         Genmask         Flags   MSS Window  irtt  Iface
> 9.2.72.0        0.0.0.0         255.255.255.128 U         0 0          0 xen-br0
> 169.254.0.0     0.0.0.0         255.255.0.0     U         0 0          0 eth0
> 0.0.0.0         9.2.72.1        0.0.0.0         UG        0 0          0 xen-br0
>
> Do you have any idea what is happening? And if you need more information
> please suggest me.
>
> Thank you very much!
>
> Regards,
> Wei Huang
>
> 774 Dreese Lab, 2015 Neil Ave,
> Dept. of Computer Science and Engineering
> Ohio State University
> OH 43210
> Tel: (614)292-8501
>
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>

Re: veth0 problems and network-route

[Keir Fraser]

On 29 Jul 2005, at 18:49, wei huang wrote:

> Do you have any back port patch to the version vnet which doesn't use
> veth0? It might help in my case. PS: I need to run program with some 
> new
> features in unstable, so I have to stick to the unstable versoin.

I think you can remove loopback.o from 
linux/drivers/xen/netback/Makefile, and that will give you old-style 
behaviour.

  -- Keir

Re: veth0 problems and network-route

[wei huang]

Hi,

Many thanks to your reply. But there has to be something else different
with xen-2.0. I didn't compile the loopback as you suggested. Now
it's close to work until the DomU boots.

The strange thing is that domU and dom0 both can talk to the whole world
but they can NOT talk to each other. They can even ping each other but
cannot ssh. (I use the same disk image as I use for xen-2.0 so I don't
think there is some forwarding policy problem, etc. And actually I turned
off all the firewall, security things)

Here is the route table of dom0 after domU boots:
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
9.2.72.0        0.0.0.0         255.255.255.128 U         0 0       0  xen-br0
169.254.0.0     0.0.0.0         255.255.0.0     U         0 0       0  xen-br0
0.0.0.0         9.2.72.1        0.0.0.0         UG        0 0       0  xen-br0

And ifconfig of dom0:
eth0      Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
          inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
vif1.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
xen-br0   Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
          inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128

ifconfig of domU:
eth0      Link encap:Ethernet  HWaddr AA:00:00:00:00:1ame:0
          inet addr:9.2.72.101  Bcast:9.2.72.127  Mask:255.255.255.128
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.X packets:18 errors:0 dropped:0

It's almost working but it seems that something is preventing those two OS
on the same machine talking to each other.

Do you have any ideas? Thank you very much!

Regards,
Wei Huang

> On 29 Jul 2005, at 18:49, wei huang wrote:
>
> > Do you have any back port patch to the version vnet which doesn't use
> > veth0? It might help in my case. PS: I need to run program with some
> > new
> > features in unstable, so I have to stick to the unstable versoin.
>
> I think you can remove loopback.o from
> linux/drivers/xen/netback/Makefile, and that will give you old-style
> behaviour.
>
>   -- Keir
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>

Re: veth0 problems and network-route

[wei huang]

Some more info:
I cannot use xencons to connect to domU. This is output of error msg:
************ REMOTE CONSOLE: CTRL-] TO QUIT ********
Cannot connect to console 0 on domain 1

************ REMOTE CONSOLE EXITED *****************
[root@lab112 ~]# xencons localhost 9601
Traceback (most recent call last):
  File "/usr/bin/xencons", line 92, in ?
    connect(str(sys.argv[1]),int(sys.argv[2]))
  File "/usr/bin/xencons", line 62, in connect
    sock.connect((host,port))
  File "<string>", line 1, in connect
socket.error: (111, 'Connection refused')


Regards,
Wei Huang

774 Dreese Lab, 2015 Neil Ave,
Dept. of Computer Science and Engineering
Ohio State University
OH 43210
Tel: (614)292-8501


On Fri, 29 Jul 2005, wei huang wrote:

> Hi,
>
> Many thanks to your reply. But there has to be something else different
> with xen-2.0. I didn't compile the loopback as you suggested. Now
> it's close to work until the DomU boots.
>
> The strange thing is that domU and dom0 both can talk to the whole world
> but they can NOT talk to each other. They can even ping each other but
> cannot ssh. (I use the same disk image as I use for xen-2.0 so I don't
> think there is some forwarding policy problem, etc. And actually I turned
> off all the firewall, security things)
>
> Here is the route table of dom0 after domU boots:
> Kernel IP routing table
> Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
> 9.2.72.0        0.0.0.0         255.255.255.128 U         0 0       0  xen-br0
> 169.254.0.0     0.0.0.0         255.255.0.0     U         0 0       0  xen-br0
> 0.0.0.0         9.2.72.1        0.0.0.0         UG        0 0       0  xen-br0
>
> And ifconfig of dom0:
> eth0      Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
>           inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
> lo        Link encap:Local Loopback
>           inet addr:127.0.0.1  Mask:255.0.0.0
> vif1.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
>           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
> xen-br0   Link encap:Ethernet  HWaddr 00:0D:60:9A:15:60
>           inet addr:9.2.72.112  Bcast:9.2.72.127  Mask:255.255.255.128
>
> ifconfig of domU:
> eth0      Link encap:Ethernet  HWaddr AA:00:00:00:00:1ame:0
>           inet addr:9.2.72.101  Bcast:9.2.72.127  Mask:255.255.255.128
> lo        Link encap:Local Loopback
>           inet addr:127.0.0.1  Mask:255.0.X packets:18 errors:0 dropped:0
>
> It's almost working but it seems that something is preventing those two OS
> on the same machine talking to each other.
>
> Do you have any ideas? Thank you very much!
>
> Regards,
> Wei Huang
>
> > On 29 Jul 2005, at 18:49, wei huang wrote:
> >
> > > Do you have any back port patch to the version vnet which doesn't use
> > > veth0? It might help in my case. PS: I need to run program with some
> > > new
> > > features in unstable, so I have to stick to the unstable versoin.
> >
> > I think you can remove loopback.o from
> > linux/drivers/xen/netback/Makefile, and that will give you old-style
> > behaviour.
> >
> >   -- Keir
> >
> >
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@lists.xensource.com
> > http://lists.xensource.com/xen-devel
> >
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>