Eric Paris [off-list ref] wrote:

[...]  Bring back up the interface connected to
eth1.  At this point we have a "valid" connection since eth1 can talk to
one of the arp targets.  But we are only sending arp requests on eth0
(verify with tcpdump)

	The trick is to have a situation with a partitioned network and
a failure such that the device still has link, but does not respond to
the ARP queries.  That's not an unreasonable failure if there's a switch
in each path to the arp_ip_target peers (which is how I set it up
locally).

The patch below has been tested by me and appears to fix the problem.
All of the failover tests I performed seem to work including pulling
cables and stopping responses from the arp_ip_target entries.  

	The patch looks good to me, also (although I made the change by
hand instead of via patch).

	-J

---
	-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com


Signed-off-by: Jay Vosburgh <redacted>

--- linux-2.6.11/drivers/net/bonding/bond_main.c.orig   2005-05-12 12:22:52.000000000 -0400
+++ linux-2.6.11/drivers/net/bonding/bond_main.c        2005-05-12 15:13:53.000000000 -0400

@@ -3046,7 +3046,7 @@ static void bond_activebackup_arp_mon(st
                        bond_set_slave_inactive_flags(bond->current_arp_slave);

                        /* search for next candidate */
-                       bond_for_each_slave_from(bond, slave, i, bond->current_arp_slave) {
+                       bond_for_each_slave_from(bond, slave, i, bond->current_arp_slave->next) {
                                if (IS_UP(slave->dev)) {
                                        slave->link = BOND_LINK_BACK;
                                        bond_set_slave_active_flags(slave);