Things To Look For In Host-Based Intrusion Prevention
Layered security is a widely accepted principle of computer and network security (see In Depth Security). The basic premise is that it takes multiple layers of defense to protect against the wide variety of attacks and threats. Not only can one product or technique not protect against every possible threat, therefore requiring different products for different threats, but having multiple lines of defense will hopefully allow one product to catch things that may have slipped past the outer defenses.
There are plenty of applications and devices you can use for the different layers- antivirus software, firewalls, IDS (Intrusion Detection Systems) and more. Each has a slightly different function and protects from a different set of attacks in a different way.
One of the newer technologies is the IPS- Intrusion Prevention System. An IPS is somewhat like combining an IDS with a firewall. A typical IDS will log or alert you to suspicious traffic, but the response is left to you. An IPS has policies and rules that it compares network traffic to. If any traffic violates the policies and rules the IPS can be configured to respond rather than simply alerting you. Typical responses might be to block all traffic from the source IP address or to block incoming traffic on that port to proactively protect the computer or network.
There are network-based intrusion prevention systems (NIPS) and there are host-based intrusion prevention systems (HIPS). While it can be more expensive to implement HIPS- especially in a large, enterprise environment, I recommend host-based security wherever possible. Stopping intrusions and infections at the individual workstation level can be much more effective at blocking, or at least containing, threats. With that in mind, here is a list of things to look for in a HIPS solution for your network:
- Doesn’t Rely On Signatures: Signatures- or unique characteristics of known threats- are one of the primary means used by software like antivirus and intrusion detection (IDS).The downfall of signatures is that they are reactive. A signature can’t be developed until after a threat exists and you could potentially get attacked before the signature is created. Your HIPS solution should use signature-based detection along with anomaly-based detection which establishes a baseline of what “normal” network activity looks like on your machine and will respond to any traffic that appears unusual. For example, if your computer never uses FTP and suddenly some threat tries to open an FTP connection from your computer, the HIPS would detect this as anomalous activity.
- Works With Your Configuration: Some HIPS solutions may be restrictive in terms of what programs or processes they are able to monitor and protect. You should try to find a HIPS that is capable of handling commercial packages off the shelf as well as any home-grown custom applications you may be using. If you don’t use custom applications or don’t consider this a significant problem for your environment, at least ensure that your HIPS solution protects the programs and processes you do run.
- Allows You To Create Policies: Most HIPS solutions come with a pretty comprehensive set of pre-defined policies and vendors will typically offer updates or release new policies to provide a specific response for new threats or attacks. However, it is important that you have the ability to create your own policies in the event that you have a unique threat that the vendor doesn’t account for or when a new threat is exploding and you need a policy to defend your system before the vendor has time to release an update. You need to make sure the product you use not only has the ability for you to create policies, but that policy creation is simple enough for you to understand without weeks of training or expert programming skills.
- Provides Central Reporting and Administration: While we are talking about host-based protection for individual servers or workstations, HIPS and NIPS solutions are relatively expensive and outside of the realm of a typical home user. So, even when talking about HIPS you probably need to consider it from the standpoint of deploying HIPS on possibly hundreds of desktops and servers across a network. While it is nice to have protection at the individual desktop level, administering hundreds of individual systems, or trying to create a consolidated report can be nearly impossible without a good central reporting and administering function. When selecting a product, ensure that it has centralized reporting and administration to allow you to deploy new policies to all machines or to create reports from all machines from one location.
There are a few other things you need to keep in mind. First, HIPS and NIPS are not a “silver bullet” for security. They can be a great addition to a solid, layered defense including firewalls and antivirus applications among other things, but should not try to replace existing technologies.
Secondly, the initial implementation of a HIPS solution can be painstaking. Configuring the anomaly-based detection often requires a good deal of “hand-holding” to help the application understand what is “normal” traffic and what is not. You may experience a number of false positives or missed negatives while you work to establish the baseline of what defines “normal” traffic for your machine.
Lastly, companies generally make purchases based on what they can do for the company. Standard accounting practice suggests that this be measured based on the return on investment, or ROI. Accountants want to understand if they invest a sum of money in a new product or technology, how long will it take for the product or technology to pay for itself.
Unfortunately, network and computer security products don’t generally fit this mold. Security works on more of a reverse-ROI. If the security product or technology works as designed the network will remain safe- but there will be no “profit” to measure an ROI from. You have to look at the reverse though and consider how much the company could lose if the product or technology were not in place. How much money would have to be spent on rebuilding servers, recovering data, the time and resources of dedicating technical personnel to clean up after an attack, etc? If not having the product may potentially result in losing significantly more money than the product or technology costs to implement, then perhaps it makes sense to do so.