mirror of
https://git.FreeBSD.org/ports.git
synced 2024-12-07 02:06:57 +00:00
24 lines
1.3 KiB
Plaintext
24 lines
1.3 KiB
Plaintext
|
This is a tool that uses ARP poisoning to have a scenario
|
||
|
|
like this: we have a LAN and we want offer connectivity to every-
|
||
|
|
one coming here with his laptop for example. It could happen that
|
||
|
|
our customer has his network parameters already configured to
|
||
|
|
work correctly in his own LAN, but not working here. We can have
|
||
|
|
then this scenario:
|
||
|
|
Customer's host (10.0.0.2/8 and default gateway set to 10.0.0.1)
|
||
|
|
Our LAN (192.168.0.0/24 with real gateway 192.168.0.254).
|
||
|
|
All that we want is that our customer plugs his laptop and joins
|
||
|
|
the internet without changing nothing of his network parameters.
|
||
|
|
Here comes this tool installed in my real gw(192.168.0.254) It's
|
||
|
|
a sort of sniffer, because it sniffs broadcast ARP requests for
|
||
|
|
the gateway and answers that the gateway is itself In our example
|
||
|
|
our customer's laptop sends this request: arp who-has 10.0.0.1
|
||
|
|
tell 10.0.0.2 Now our gateway does the following: 1) Sends back
|
||
|
|
this reply to 10.0.0.2: arp reply 10.0.0.1 is-at his_mac_address
|
||
|
|
2)Create the alias 10.0.0.254 (ARP is not routable so we need one
|
||
|
|
alias for each subnet that is not our one) 3)Sends itself an ARP
|
||
|
|
reply to refresh his ARP cache
|
||
|
|
It is different from proxy arp for two reasons: first it runs in
|
||
|
|
user space, then in this case we can plug machines belonging to
|
||
|
|
whatever subnet, while proxy arp is used in the case of only two
|
||
|
|
different ones.
|