LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: Arp a problem?

To: Ard van Breemen <ard@xxxxxxxxxxxxxxx>
Subject: Re: Arp a problem?
Cc: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
From: Julian Anastasov <ja@xxxxxx>
Date: Thu, 2 Nov 2000 22:23:14 +0000 (GMT)
        Hello,

On Thu, 2 Nov 2000, Ard van Breemen wrote:

> >     Yes, your router is "good" just like a Linux router: the
> > ARP REPLY is send to the MAC address of the real server and not to
> > the VIP (192.168.16.2).
> Now you got me really curious... In what cases (or better: which
> level(2/)3 routers) does the router not comply to the arp rfc?

        To be precise the router's behavior is expected, i.e. the
the reply is send to the source hardware address.

> I've seen the right arp-behaviour before, and with different operating
> systems: ncr-unix, and solaris. Sometime and company ago some mix up of
> the switches put some machines "physically" in the same net as me, but
> another subnet. I just added the routeing for the machines to my system,
> and ping -R ed them. And yes: broadcast arp request for the system,
> unicast arp reply directed to me from the system. Icmp echo request
> send to the system, and icmp-echo-reply received from the router :).
> It seemed the icmp-echo travelled from me, directly to the system, then
> to the systems default gateway, and a few routers more back to me :).

        Yes, may be Linux is unique, i.e. every local address can
be announced. I had a chance to discuss this behavior on the linux-kernel
mail list but I still don't know why we can't always use the preferred
source address in the ARP requests. Currently, the other hosts are
usually confused from such requests with unknown addresses.

> --
> Ard van Breemen, T(elegraaf)E(lektronische)M(edia)
> http://www.faqs.org/rfcs/rfc1855.html
> **THIS E-MAIL MESSAGE IS VIRUS FREE BY COMPLYING TO THE ASCII STANDARD**


Regards

--
Julian Anastasov <ja@xxxxxx>



<Prev in Thread] Current Thread [Next in Thread>