All Cisco-Network Study Notes: Understanding the Blocking Process

Understanding the Blocking Process

Threats to our networks never sleep or wait until Monday to become a burden. While there are many mitigating tools available to the security administrator such as firewalls, password security, encryption standards, and even complex networks with private IP addresses, malicious traffic still seems to find its way into the network. Hence, we have the need for network intrusion detection systems, or NIDSs, to find these intruders and make the administrator's aware of the threats to their network and data.

However, Security Administrators and their teams may not always be available and nobody likes to be awoken in the middle of sleep, or called while vacationing in Bermuda, to find out their network has just been a victim of a Denial-of-Service attack, web defacement, or worse. This is where IP blocking can come in handy. Instead of just watching, logging, and sending alerts as malicious activity happens on the network, blocking can be put in place to dynamically stop intruders and allow ample time for someone to take the needed corrective actions.

If blocking is configured correctly, it can also stop this traffic from using any other possible entrances to the same network, say from a redundant connection with a different ISP, and keep the unwanted traffic out however long the Security Administrator has configured it to do so. We can see in Figure 8.1 that the IDS sensor can send the Access Control List "199" to both Internet bordering routers, thus effectively keeping the attacker from entering the network at either point. Now the Security Administrator can rest easier and handle the problem when he is ready to handle the problem, not work according to the attacker's schedule.

Figure 8.1: Blocking Multiple Network Entry Points

What Is Blocking?

So what actually is IP blocking? IP blocking is very much as it sounds, when a Cisco Secure IDS sensor detects malicious behavior, IP blocking is implemented to deny further network traffic entering our network from the identified source host IP address. What's more, we can configure what specific traffic patterns, or signatures, we want the intrusion detection system (IDS) to watch for and manage according to an administrator-assigned severity level.

IP blocking makes use of Access Control Lists (ACLs) to stop the suspect traffic from entering the network and will block the network traffic until the block is either manually removed or it exceeds it's predetermined blocking duration.

Access Control Lists

At this point, it is important to have a thorough understanding of Cisco's Access Control Lists (ACLs), sometimes simply referred to as access-lists. There are two main types of ACLs we will be concerned with in this chapter. These are the Standard Access Control List and the Extended Access Control List. A number that is assigned to them can quickly identify these two types of ACLs. Two steps must be followed when creating an access-list. The first step is to create the list itself and the second is to apply the list to an interface.

Step 1: Standard Access-Lists

These ACLs are used in situations when we only want to block a source IP host address or an entire network IP address from accessing a certain network. This is specifically working with the IP addresses and not giving any mind to other information, such as port number or protocol. The number assigned to a standard access-list will be in the range of 1–99, and an expanded range of 1300–1999. Here is the basic format of a standard access-list:

access-list access-list-number [permit|deny] [source-ip-address] [source-
   wildcard]

Here is an example:

access-list 19 permit 172.16.2.0 0.0.0.255

The access list components are as follows:

access-list The command used to initiate the creation of an access-list
access-list-number A number within the range previously specified for the type of access-list being created.
[permit|deny] This tells the network device to either "permit" the traffic to pass the interface or to "deny" the traffic from passing and drop the traffic.
source-ip-address Indicates the particular host IP or network IP address to be checked.
source-wildcard Used to indicate an actual host, subnet, or network. This is broken down into comparing subnet masks and the ip address. A "0" indicates the number in its respective octet matches perfectly. Any "1s" indicate any number can be used. For instance, in the example, for any ip to be matched by the access-list, 172.16.2 must be the first part of the IP address. The last octet can be anything from 0–255. The example "permits" any traffic from the IP range of 172.16.2.1–172.16.2.254 to pass the network device's interface.

Note
More on wildcard masks and their usage can be found in various articles at www.cisco.com.

Let's look at an example to better understand how this configuration works on a network referring to Figure 8.2.

Figure 8.2: Implementing a Standard Access-List

Let's say Client1 on the Network XYZ needs access to resources located on the Network ABC. By using an access-list we can provide this admission and still protect the network from others trying to enter. Here is the configuration shown as an excerpt from a show run command:

Router1
!
interface Serial0
ip address 192.168.0.1 255.255.255.252
ip access-group 10 in
!
access-list 10 permit 10.0.0.10
!

Notice the access-list has been applied to the Serial0 interface on Router1 with the ip access-group 10 in statement. This configuration will permit only incoming traffic from the address 10.0.0.10, and allow it access to the entire ABC network.

If it were desired to block Client1 on XYZ and allow the rest of the network, we would use the configuration:

Router1
!
interface Serial0
ip address 192.168.0.1 255.255.255.252
ip access-group 10 in
!
access-list 10 deny 10.0.0.10
access-list 10 permit any
!

This configuration will allow all users to access every area of the network except Client1. Keep in mind that access-lists are read from top to bottom. When a match is found, that match will be used and the router stops comparing the packets to the access-list. Therefore, once the router reads the "permit any" statement, it will stop reading and simply allow "any" address through. Also, the most heavily matched items should be placed at the beginning of the access-list to help ease the CPU load.

Note

Not shown in the configuration is an implicit "deny all" statement. This does not need to be manually edited because it is automatically applied. This is a wonderful security feature, but if you are not aware of its existence, it can lock out all critical systems attempting to access resources on the network. As in the preceding configuration, only the host 10.0.0.1 will be allowed to traverse Router1. All other traffic attempting to enter this interface will be dropped.

Step 2: Extended Access-Lists

These ACLs give us much more depth in how to control network traffic. Extended access-lists can be configured to check port number, protocol, and the destination address as well as the source address. The number assigned to an extended access-list is in the range of 100–199, and an expanded range of 2000–2699. Here is the basic format of an extended access-list:

access-list access-list-number [permit|deny] protocol source ip address 
    source-wildcard destination destination-wildcard [operator]

Here is an example:

access-list 190 deny TCP 172.16.1.0 0.0.0.255 172.16.2.0 0.0.0.255 neq 23

The extended access-list is different than the standard ACL in the following ways:

access-list-number This is a number within the range previously specified for the type of access-list being created. In this case, it is an extended access-list, as indicated by the 190.
protocol This allows us to filter based on a particular protocol—typically, IP for filtering by IP address or TCP for filtering by protocol.
destination This is the destination IP address a particular packet is trying to reach. In the example, it is the network address of 172.16.2.0.
destination-wildcard This works as the source-wildcard to indicate the host IP address or range of IP addresses trying to be reached. This saves the time of having to type lines for each IP address within a particular subnet.
Operator This can be used to indicate a port number when filtering by protocol, such as TCP or UDP. There are four options that can be used with this feature.
- eq Equals—when we know exactly what port needs to be monitored
- gt Greater than—allows us to specify a particular range over a particular port number
- lt Less than—allows us to specify a particular range lower than a particular port number
- neq Not equal—allows us to assert the access-list to all but on port

Let's now see how this works on a network. Refer to Figure 8.3.

Figure 8.3: Implementing an Extended Access-List

Looking at this network example, imagine that all the clients need access to the e-mail server and only Client3 needs Telnet access to the payroll server. Our configuration would look similar to this:

Router1
!
interface Ethernet1
ip address 172.16.2.1 255.255.255.0
ip access-group 110 out
!
access-list 110 permit tcp 172.16.1.30 any eq telnet
access-list 110 deny tcp any 172.16.2.10
access-list 110 permit tcp 172.16.1.0 0.0.0.255 any
!

This access-list configuration first allows Client3 to access to the Demilitarized Zone (DMZ) and specifically access the payroll server using Telnet (indicated by eq). The next line in the access-list is to deny anyone to access the payroll server. Notice this comes after the line that allows Client3 access to the same server. Remember, when the router finds the first match, it will stop reading and take action. The third line allows anyone from the internal network 172.16.1.0 to access any resource and any protocol, and so on, not already denied, on the DMZ. If a match is still not found, the implicit deny will drop all remaining network traffic trying to access the DMZ.

One condition we used in the preceding configuration had to do with whether we applied the access-list to an internal or external interface and the particular direction of the traffic flow. In the last example, we can see the ACL was applied to an external interface, in relation to the internal network. We also have the command "out" used. This means that traffic trying to leave the router on the Ethernet1 interface will be checked. The first example, Figure 8.1, used an ACL on the Serial0 interface, applying the in command. This prevents the network traffic specified in the ACL from entering the router. The traffic is then dropped right at the front door.

Note

Only one access-list can be applied to an interface/direction at a time. If another ACL is applied to the same interface/direction, the original ACL will be nullified.

IP blocking relies on a system that will compare inbound and/or outbound IP datagrams on a particular interface to a list of signatures and trigger an alarm if a match is made. The alarm will indicate to the governing sensor that there is a threat. At this point, the sensor will create a new ACL and distribute that ACL to the network device providing access to the network in jeopardy. When the new ACL reaches the network device, it will replace the resident ACL with the newly arrived one, thus dynamically updating the network device. This process is referred to as Device Management.

Device Management

The process of device management includes the construction of a new ACL and applying that ACL to the appropriate network device interface thus stopping, or blocking, the network breach immediately. As mentioned earlier, the block will remain in effect until either it has been removed manually or the blocking duration has expired. Device management relies on a Telnet connection, via its command and control interface, to the network device in question. For the sensor to successfully Telnetted to the network device, it requires the following:

There must be a route to the network device within the device management list on the sensor or the sensor must reside within the same subnet as the network device.
The network device must be configured to accept Telnet connections by configuring the vty lines with a login password and, if used, a Telnet username, and allow secure configuration changes by setting an enable password.

It is important to keep in mind that the IDS sensor will be contacting the router with an enable-privileged password. This could foster an insecure environment if this password were to get out. When possible, it is recommended to keep the Telnet sessions isolated to a "Telnet only" network. Another option is to apply ACLs to lock down all Telnet access to the network device except from approved systems and/or users.

Device management uses extended ACLs to provide its blocking functionality. The ACLs are 198 and 199. The first ACL to be used by Device Management is 199. When a change is demanded via an alarm from the sensor, Device Management will create a new ACL—that's right, 198—and send it to the rescue.