[Xen-users] DHCP problems with IP alias

[Lin Salisbury <lin.salisbury@xxxxxxxxxxxxxx>]

Hi all,

We are experiencing a strange problem with DHCP+Xen running Debian  
Etch and xen-3.0.3-1.  We have a setup where we use aliased (such as  
eth0:1) IP addresses to provide 'well-known' IP's on a network where  
the 'primary' interface (eth0 in this case) is setup as DHCP.  An  
example /etc/network/interfaces from our setup:
auto lo
iface lo inet loopback

auto eth0
iface eth0 inet dhcp

iface eth0:1 inet static
  address x.x.x.y
  netmask 255.255.255.0
  gateway x.x.x.y1

We ifup eth0:1 when we want the static IP address available, and  
ifdown when we don't.  In our case the DHCP server is running on the  
Image that is affected, all other images with this setup act ok.  But  
in this case when eth0 tries to reacquire its IP address from DHCP  
(which is running locally on this image) it fails with the udp  
problem and the eth0:1 IP address goes away.  Here is the output of  
ethtool -k eth0:
Offload parameters for eth0:
Cannot get device rx csum settings: Operation not supported
Cannot get device udp large send offload settings: Operation not  
supported
rx-checksumming: off
tx-checksumming: off
scatter-gather: off
tcp segmentation offload: off
udp fragmentation offload: off
generic segmentation offload: off

The workaround we have is to actually create a separate VIF for the  
static IP address and not have it aliased to eth0, and everything  
seems ok.  I haven't tried Xen 3.1 yet, so perhaps this issue is  
fixed there.  But does anyone else have similar problems perhaps with  
heartbeat or other HA tools?
TIA,
Lin


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