Re: RE: Discovering network subnets

[<chad () mr-lew com>]

  Let's try to clear this up a bit...

If you have a network mask of 255.255.255.0 (24 bits), then 
the .0 address is your Network ID and can NOT be used as a 
host address. The .255 address will be your broadcast 
address and also can not be used as a host address. This 
makes .1 thru .254 as valid host addresses.

192.168.1.0/24 - Network ID (All Host bits are 0)
192.168.1.1-254 - Host Addresses
192.168.1.255 - Broadcast Address (All Host bits are 1)

   You may see some responses from a Cisco router, not 
actually a host address, but it may respond to some of the 
ICMP probes. I am sure some other systems performing routing 
functions may do the same in some circumstances, but in my 
experience they do not answer up with SYN/ACK or RST/ACK for 
port scans.

Now if you have a network mask of 255.255.254.0 (23 bits) or 
anything less than 24 bits for that matter, a .0 can and IS 
a valid host address.

In the example of 10.0.0.0/23 the Network ID would be the 
first 23 bits, making it 10.0.0.0. The first available host 
would have the 32nd bit turned on, making it 10.0.0.1. The 
last available host would have bits 24 thru 31 turned on, 
with the 32nd bit turned off making it 10.0.1.254. This 
would include 10.0.1.0 as a VALID host address.

Where the confusion comes from is crossing the bit boundary. 
You need to look at it in binary to see how it is just the 
next host when going from 10.0.0.255 to 10.0.1.0. Hopefully 
this diagram can help (and won't get butchered in 
delivery) ;)

      1                   1
      2 6 3 1             2 6 3 1
10.0. 8 4 2 6 8 4 2 | 1 . 8 4 2 6 8 4 2 1
----------------------------------------
   NETWORK PORTION  |  HOST PORTION
--.-. 0 0 0 0 0 0 0 | 0 . 0 0 0 0 0 0 0 0   Network ID
--.-. 0 0 0 0 0 0 0 | 0 . 0 0 0 0 0 0 0 1   1st Available 
Host
--.-. 0 0 0 0 0 0 0 | 0 . 1 1 1 1 1 1 1 1   Valid Host
--.-. 0 0 0 0 0 0 0 | 1 . 0 0 0 0 0 0 0 0   Valid Host
--.-. 0 0 0 0 0 0 0 | 1 . 1 1 1 1 1 1 1 0   Last Available 
Host
--.-. 0 0 0 0 0 0 0 | 1 . 1 1 1 1 1 1 1 1   Broadcast Address

Also, Classful addressing was not done away with by CIDR. 
CIDR granted us the ability to better use and 
identify/aggregate our networks. Classful addressing is 
still used today in numerous places (RIPv1 for example), but 
when possible classful addressing is normally preferred.

It was pointed out that RFC 3021 outlines the specific use 
of a 31 bit mask, which would break this model if used. It 
must be pointed out that the RFC outlines this use on point-
to-point links only. Considering that the system 
was "reported" as having ports 68/tcp, 723/tcp and 6000/tcp 
open, I would be inclined to rule against it being on a 
point-to-point link.

68/tcp - ??? Could this be some TCP based BOOTP server...
723/tcp - OpenMosix File System (curious)
6000/tcp - Probably X Windows

I would try to gather some more O/S fingerprinting 
information by generating more ICMP and TCP responses 
(SYN/ACK and RST/ACK) with hping2 and then gather the 
responses with a sniffer to try and use p0f to get a better 
picture.

I would be curious to hear what the final findings are...


------------------------------------------------------------------------------
FREE WHITE PAPER - Wireless LAN Security: What Hackers Know That You Don't

Learn the hacker's secrets that compromise wireless LANs. Secure your
WLAN by understanding these threats, available hacking tools and proven
countermeasures. Defend your WLAN against man-in-the-Middle attacks and
session hijacking, denial-of-service, rogue access points, identity
thefts and MAC spoofing. Request your complimentary white paper at:

http://www.securityfocus.com/sponsor/AirDefense_pen-test_050801
-------------------------------------------------------------------------------