Security Basics mailing list archives

RE: What could this icmp mean?


From: "David Gillett" <gillettdavid () fhda edu>
Date: Thu, 24 Feb 2005 14:00:49 -0800

1.  Your devices on the 10.30.1.x network have 10.30.1.254 as their
    default gateway, but traffic to 10.30.0.x needs to go by 10.30.1.1
    instead.  10.30.1.254 is attempting to inform 10.30.1.16 that if
    it wants to send traffic to 10.30.0.4, it could save a hop by
    sending those packets directly to 10.30.1.1 instead of relying
    on 10.30.1.254 to forward them.

    Since most clients don't know what to do with ICMP redirects, and
    will just ignore them, it's common to turn them off at the router
    interface.

2.  ICMP packets carry, as payload, a portion of the packet that triggered
    the ICMP.  It's no surprise that the checksum contained within this
    partial quotation is a checksum for the full packet, and not just the
    quoted portion.  It would be a near-miracle if these computed to the
    same value.

David Gillett


-----Original Message-----
From: Tomas [mailto:s.tomas () gmail com]
Sent: Tuesday, February 22, 2005 5:11 AM
To: security-basics () securityfocus com
Subject: What could this icmp mean?


Hello list,

We have networks (10.30.0.0/24 and 10.30.1.0/24) connected 
trough VPN and
one internet line. The gateways for VPN are 10.30.0.1 from 
one side and
10.30.1.1 from the other, and 10.30.1.254 for internet (for 
both networks).

I've launched tcpdump today on my internet firewall's 
internal interface
(10.30.1.254) and I found this:

10.30.1.254 > 10.30.1.16: icmp: redirect 10.30.0.4 to host 
10.30.1.1 for
10.30.1.16.445 > 10.30.0.4.1959: [|tcp] (DF) (ttl 127, id  
7691, bad cksum
c76d! differs by 100) (ttl 255, id 23807)

I'm a bit confused, what could this icmp mean? First of all, 
I'm sure that
neither of these hosts (10.30.1.254, 10.30.1.16, 10.30.0.4) 
are sending any
icmp requests (I'm not sure about 10.30.1.1; it's not in my 
control). And
the second of all, why the checksum is bad?



Current thread: