What if target is behind NAT
Configuring hosts to be Internet- and LAN-enabled
Each computer on
the Internet has to have TCP/IP software, and also Link Layer software
(Ethernet, wireless, etc.) and appropriate NIC.
The parameters
needed to operate on Internet are: MAC address, IP address, subnet mask,
default gateway, and DNS server.
MAC
address comes with the NIC automatically.
Software
and NIC must be configured with IP address, subnet mask, default gateway, and DNS
server. This is duty of IT personnel.
Configuration can
be manual or automated.
If a network device
is configured to have a static IP, then the IP, the subnet mask, DSN server and
default gateway must be manually entered into TCP/IP configuration menu. Those
numbers will not change.
If a network device
is configured to use DHCP, then it will automatically get its own IP address,
subnet mask, default gateway IP, and DNS server IP from DHCP server. It’s IP
address will be dynamically assigned, i.e. it will not be a permanent number.
Network Address Translation (NAT) protocol
In IPv4, there are not enough IP addresses, and it is
necessary to invent some “smart” tricks to make one address act as many. Enter NAT:
a router acts as a proxy for IP addresses, mapping many private addresses to
one public address. NAT is used so that one IP unique IP address can be
“stretched out” to represent many computers, therefore saving IP address space.
In addition, NAT is useful to ensure security: computers behind NAT do have
private addresses and are not accessible from the outside network.
Besides NAT, there are also:
NAPT: Network Address Port Translation
PAT: port address translation. Used for SOHO. ISP assigns a “real” i.e.
public IP address to the router. Router assigns unique port numbers to the home
computers.
Typical configuration of SOHO
networks:
|
|||
Clients and default gateway are all configured to get
information via DHCP.
However: clients get it from the gateway, i.e. the gateway
is their DHCP server.
Gateway gets it from the ISP DHCP server.
ISP’s DHCP server issues unique IP address to the default
gateway. ** This address will be automatically downloaded by all clients when
they connect into the network. **
If we talk about a larger organization and not SOHO, then
the organization itself will “book” one “real” i.e. static IP address for the
default gateway.
Of course, this discussion pertains to IPv4 addressing. In
case of IPv6, there is no need for NAT. There are enough IPv6 addresses and
therefore, each device on Internet can have its own public address. That seems
too insecure to sys admins who are used to NAT. NAT does not provide solid
security by any means, because computers on the private network behind NAT
usually can reach outside to the public network, and if an attacker wants to,
attacker can reach the private network. However, it does require more work and
effort than not having the NAT and having computers on public network.
Therefore, NAT is generally considered an additional layer of security and thus
people are not readily switching to using IPv6 and foregoing NAT>
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home