The Packet Filtering Firewall is one of the most basic firewalls. The first step in protecting internal users from the external network threats is to implement this type of security. The first ever firewalls used were of packet filtering type only. As the trends of network threats started changing, so did the firewall building strategies. Most of the routers have packet filtering built-in, but the problem with the routers is that, they are difficult to configure and don’t provide extensive logs of the incidents. In my previous firewall tutorials I talked about firewall policies and few other things. That information is also used while designing such firewalls.
To start with the network security, the packet filtering firewalls are the way to go. This functinality is still the main aim of most of the commercial and non-commercial firewalls. As you know by the definition and the purpose of the firewall, the firewall is the first destination for the traffic coming to your internal network. So, anything which comes to your internal network, passes through the firewall. Of course, reverse is also true. Any outgoing traffic will also pass through the firewall before leaving your network completely. This is the reason that sometimes this type of firewall filter is also called screening routers.
Types of Packet Filtering
Packet filtering firewall allows only those packets to pass, which are allowed as per your firewall policy. Each packet passing through is inspected and then the firewall decides to pass it or not. The packet filtering can be divided into two parts:
- Stateless packet filtering.
- Stateful packet filtering.
The data travels through the internet in the form of packets. Each packet has a header which provides the information about the packet, its source and destination etc. The packet filtering firewalls inpects these packets to allow or deny them. The information may or may not be remembered by the firewall.
Stateless Packet Filtering
If the information about the passing packets is not remembered by the firewall, then this type of filtering is called stateless packet filtering. This type of firewalls are not smart enough and can be fooled very easily by the hackers. These are especially dangerous for UDP type of data packets. The reason is that, the allow/deny decisions are taken on packet by packet basis and these are not related to the previous allowed/denied packets.
Stateful Packet Filtering
If the firewall remembers the information about the previously passed packets, then that type of filering is stateful packet filtering. These can be termed as smart firewalls. This type of filtering is also known as Dynamic packet filtering.
What Should Be Inspected In A Packet Header
In a packet header few of the possible things which should be checked are:
- Source IP address of the packet. This is necessary because IP spoofers might have changed the source IP address to reflect the origin of packet from somewhere else, rather than reflecing the original source.
- Destination IP Address. The firewall rules should check for IP address rather than DNS names. This prevents abuse of DNS servers.
- IP Protocol ID.
- TCP/UDP port number.
- ICMP message type.
- Fragmentation flags.
- IP Options settings.
Important Features of Packet Filters
The great firewalls normally follow few specific rules upon which features are incorporated during firewall designing. Few are listed below:
- The firewall should provide good deal of logs. The more detailed are the logs, the better the protection.
- The command line syntax or GUI of firewall should be easy to create new rules and of course firewall exceptions.
- The packet filter orders should be evaluated carefully in order to make the filtering fruiteful.
At the end, a word of caution. You SHOULD NOT depend only upon packet filtering firewalls for the security of your network. Firewalls network security is not fully reliable, we need to take several other measures in order to have full network security.