Attacking BGP
Loki contains a universal BGP module, written in python. It implements the most common used BGP packet and data types and can be used to establish a connection to a BGP speaking peer. Once a connection is established, the tool starts a background thread which sends keep-alive packages to hold the connection established and the published routes valid. To publish BGP routing information the module provides built-in data types which can be merged to the appropriated update statement. Once an update statement is set up it can be send once or multiple times to the connected peer. It is possible to use kernel based MD5 authentication, as described in RFC2385. Another module makes it possible to brute force the used MD5 authentication key.An Example for Injecting IPv4 Routing Information
The peer is a Cisco 3750ME with a (pre-attack) routing table looking like this:
Loki is then used to inject IPv4 routing information:
The first step is to configuring the target IP address, the autonomous system number 2 and a hold timer of 8 seconds. Afterwards the session can be established by clicking on the “Connect” button. If Loki is able to establish the connection, a background keep alive thread is started, which sends an BGP keep alive packet every hold time / 4 seconds. The next step is to configure the BGP update message, which defines, the routing information to publish to the connected host. In the example case we build up a RFC1771 IPv4 routing BGP update packet which says we are announcing the network 192.168.233.0/24 and traffic for this network should be forwarded to the IP address 10.0.0.2 which is our attack host. In the end we send the prepared update packet out by selecting the designated host from the connection list and clicking the “Update” button.
After publishing the routing information, the router’s routing table looks like this:
So we injected a route to the network 192.168.233.0/24 which, in this case, directs all matching traffic to our (attack) host.
Comments