Vincent Guffens wrote:
> ok, I see, so both interfaces are replying to the arp request but this time, the first interface to reply, eth0, will have its
> arp reply filtered by the arp_filter as ip_route_output will indicate eth1 as the outgoing interface for a packet that has its
> source set to the ip to resolve, which has a source route policy entry to point to eth1, is it ?
Correct. Both NICs receive the arp request, only one of them actually sends a
reply.
> But then, by curiosity, is there some reasons why two different interfaces shouldn't be in the same subnet ? I was in this
> impression having configured some cisco routers which don't even accept this configuration. Is it possible on other network
> devices ? Is there any theoretical justification ?
I don't think there are any theoretical objections. I'm pretty sure windows
will let you do it, but they don't give you nearly as much control over the
routing.
Without source routing or arp filter, traffic to any of the interfaces on the
same subnet can go through any of the interfaces on that subnet--whichever arp
reply gets there last is the winner. This can be useful for quasi-redundant
type communications, but causes big problems if you're trying to monitor the
state of your network communications paths. This is where the more strict
behaviour becomes useful.
Chris
-- Chris Friesen | MailStop: 043/33/F10 Nortel Networks | work: (613) 765-0557 3500 Carling Avenue | fax: (613) 765-2986 Nepean, ON K2H 8E9 Canada | email: cfriesen@nortelnetworks.com - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Apr 23 2002 - 22:00:36 EST