[Fedora-xen] RHEL 5 - Multiple Bridges with same Hardware Address

[Ubaidul Khan]

I am writing this email in the hopes of getting some advice from Xen users.  
My
main concern is performance and stability.  Before I get any further, let me
explain the environment:

	1.      I am using redhat enterprise linux 5 with the xen kernel
		 (2.6.18-8.1.1.el5xen).  This is on a Sun X4200 with two dual core AMD
		 processors, 16Gb of RAM, a Sun StoredgeTek 3320 attached to the
		 X4200, has a 4 port NIC.

	2.     eth0 is configured as the primary interface for the Domain-0 and is
		attached to the xenbr0.  I am using network-bridge model to provide
		networking for the two guest domains.  One of the domains is running NFS
		server and needs to be configured with a private network address(on eth1).

	3.     So what I did is, I took eth1(on domain-0) and configured it on the 
private
                network.  Then I created a script called 
*custom-bridge*(this was mentioned in
		(http://www.debian-administration.org/articles/470), with the following:

     +-----------------------------------------------------------
      |   #!/bin/sh
      |   dir=$(dirname "$0")
      |   "$dir/network-bridge" "$@" vifnum=0 netdev=eth0 bridge=xenbr0
      |   "$dir/network-bridge" "$@" vifnum=3 netdev=eth1 bridge=xenbr1

		and modified xend-config.sxp in the following manner:

     +-----------------------------------------------------------
      |   # The bridge is named xenbr0, by default.  To rename the bridge, 
use
      |   #
      |   # (network-script 'network-bridge bridge=<name>')
      |   #
      |   # It is possible to use the network-bridge script in more 
complicated
      |   # scenarios, such as having two outgoing interfaces, with two 
bridges, and
      |   # two fake interfaces per guest domain.  To do things like this, 
write
      |   # yourself a wrapper script, and call network-bridge from it, as 
appropriate.
      |   #
      |   (network-script 'custom-bridge start')

	The guest domain comes up fine and I am able to talk on the private network
	from the guest domain and also from the private network to the guest 
domain.
	However, both of the bridges are configured with the same hardware address:


      xenbr0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
      xenbr1    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF

	I think this is causing the following messages to show up on logs:

Jul  9 17:15:25 kernel: peth0: received packet with  own address as source 
address
Jul  9 17:19:38 kernel: peth0: received packet with  own address as source 
address
Jul  9 17:19:38 kernel: peth0: received packet with  own address as source 
address
Jul  9 17:19:42 kernel: peth0: received packet with  own address as source 
address
Jul  9 17:30:27 kernel: peth0: received packet with  own address as source 
address
Jul  9 17:30:27 kernel: peth0: received packet with  own address as source 
address
Jul 10 09:12:19 kernel: peth0: received packet with  own address as source 
address
Jul 10 10:06:04 kernel: peth0: received packet with  own address as source 
address
Jul 10 10:06:04 kernel: peth0: received packet with  own address as source 
address
Jul 10 10:06:05 kernel: peth0: received packet with  own address as source 
address


I am also seeing performance degradation.  Any advice/suggestion is greatly
appreciated.

_________________________________________________________________
http://liveearth.msn.com