Traffic Flow

In most cases, firewalls function as a gateway between two end-points. Network traffic flows through the firewall.

  • IN, traffic coming into a firewall interface
  • OUT, traffic going out a firewall interface

Our objective pre-deployment, and during investigations, is to validate that our understanding of a given ruleset behaves as expected. Implementing a block by default ruleset, simplifies the flow validation process:

match ruleset design are:

  • passed as expected and
  • blocked otherwise.

Corporate Policies are embedded as our "ruleset". Two tools very useful in generating and monitoring traffic flow are:

Test Environment

Tests should be engineered to:

  • validate behaviour for positive rule match and
  • validate behaviour for negative rule match negative

The Test environment needs to simulate the actual gateway function for the firewall. At minimum the environment will require three hosts representing:

a. The firewall/gateway b. Hosts to one side of the firewall/gateway c. Hosts on the other side of the firewall/gateway

Writing Tests

Tests are gleaned from the proposed firewall rules, as moving betweeen rule symantics into flow semantics.

from ruleset:

pass in on $int_if inet proto tcp from <lan> to any
pass out on $ext_if

evolves to flow routes

<lan> to dmz {tcp}
<lan> to inet {tcp}

The firewall/gateways should remain essentially static during the test and various changes will be needed on the Internal and External hosts to simulate the appropriate IP Addresses to validate the rulesets

During IP Address and routing changes on the hosts, ensure route states are cleared (or the hosts restarted) to minimise errors caused by stagnant routing information.

netcat - generating controlled traffic packets

nc(1) - netcat utility, push and listen to TCP/UDP connections on interfaces

Generating and Recieving traffic using netcat provides an environment for controlled traffic flow that can be monitored, validated.

The nc (or netcat) utility is used for just about anything under the sun involving TCP or UDP. It can open TCP connections, send UDP packets, listen on arbitrary TCP and UDP ports, do port scanning, and deal with both IPv4 and IPv6. Unlike telnet(1), nc scripts nicely, and separates error messages onto standard error instead of sending them to standard output, as telnet(1) does with some.

The simplest test of traffic flow is to run netcat on a 'client' to one side of the firewall and a recieving/ listening netcat 'server' on the opposite side of the firewall.

TCP packets

We use the destination-ip and the port-number explicitly to allow us to use aliases on the destination host (simplifies testing)

Server

nc -4kl destination-ip port-number

Client

nc -4vvvu destination-ip port-number

UDP packets

There seems to be something perculiar with using -k. When using -k be explicit in the destination ip. When destination-ip is not explicit, then it may not as work expected with -k.

Server

nc -4vvvvvklu destination-ip port-number

Client

nc -4vvvvvu destination-ip port-number

tcpdump - monitoring logged packets

PF can record network packet headers and data when the log key word is used with a rule. When a packet matches a rule with the log key word, the headers and packet body are sent to the pflog pseudo-device. Once a packet is logged to the pflog pseudo-device, The tcpdump utility can be used to print the packet's contents in real time:

sudo tcpdump -netttvvi pflog0
Dec 03 15:31:17.263703 rule 5/(match) [uid 0, pid 24100] 
  block in on em1: 192.168.18.19.1055 > 192.168.18.255.1100: udp 99 (ttl 128, id 12204, len 127)

Network NICs

A common problem while evaluating host configuration changes, is the state of routing tables. Misunderstanding about the current state of the routing table can lead to redherrings. (false redirections.)

Here's a sequence of activities for clearing the various related routing 'buffers'.

  • flush the routes
  • delete arp table
  • delete the interface eth0 configuration
sudo route flush -inet
sudo arp -ad
sudo ifconfig eth0 delete

Then you can re-enable things again.

  • configure the interface
  • configure the route
sudo ifconfig eth0 __.__.__.__  netmask __.__.__.__
sudo route add default __.__.__.__