When my wife an I were a young couple looking for a place in the hills of Berkeley, we came across an ad for an apartment with “Views of the Golden Gate Bridge”. The price was a bit over our budget and the neighborhood was less than thrilling, but we decided to check it out. We had both previously lived in places with bay views and we felt that the extra expense would be worth it. But after we got to the property the apartment was beyond shabby, and no place we wanted to live. What’s more, we could not find a view! We stayed for a while searching for the advertised view, and when neither of us could find it we asked the agent. She said the view was from the side of the house. As it turns out, if you either stood on the fence in the alley, or on the toilet seat of the second bathroom, and looked out the small window, you could see a sliver of the Golden Gate. The agent had not lied to us – technically there was a bridge view. But in a practical sense it did not matter. I would hardly invite company over for a glass of wine and have them stand on tiptoes atop the toilet lid for an obstructed view of the bridge.

I think about this often when I read security product marketing collateral. There are differing degrees of usefulness of security products, and while some offer the full advertised value, others are more fantasy than reality. Or require you to have a substance abuse problem – either works.

This is of course one of the things we do as analysts – figure out not only whether a product addresses a security problem, but how usefully it does so, and which of the – often many – use cases it deals with. And that is where we need to dig into the technology to validate what’s real vs. a whitewash.

One such instance occurred recently, as I was giving a presentation on how malware is the scourge of the earth, nothing solves the problem, and this vendor’s product stops it from damaging your organization. If you think my paraphrasing of the presentation sounds awkward, you are not alone. It was. But every vendor is eager to jump on the anti-malware bandwagon because it is one of the biggest problems in IT security, driving a lot of spending. But the applicability of their solution to the general use case was tenuous.

When we talk to IT practitioners about malware they express many concerns. They worry about email servers being infected and corporate email and contacts being exposed. They worry that their servers will be infected and used to send across the Internet. They worry that their servers will become bots and participate in DoS attacks. They worry that their databases will be pilfered and their information will stream out to foreign countries. They worry that their users will be phished, resulting in malware being dropped on their machines, and the malware will assume their user credentials. They even worry about malware outside their networks, infecting customer machines and generating bogus traffic. And within each of these use cases are multiple attack avenues, multiple ways to pwn a machine, and multiple ways to exfiltrate data. So we see many legitimate techniques applied to address malware, with each approach a bit better or worse suited, depending on the specifics of the infection. And this is where understanding technology comes in, as you start to see specific types of detection and prevention mechanisms which work across multiple use cases. Applicability is not black and white, but graduated.

The solutions that only apply to one instance of one use case make me cringe. As with the above reference vendor, they addressed a use case customers seem least interested in. And they provide their solution in a way that really only works in one or two instances of that use case. Technically the vendor was correct: their product does indeed address a specific type of malware in a particular scenario. But in practice it is only relevant in a remote niche of the market. That is when past and present merged: I was transported back in time to that dingy apartment. But instead of the real estate agent it was the security vendor, asking me to teeter on the toilet seat lid with them, engaging in the fantasy of a beautiful Golden Gate Bridge view.

On to the Summary:

Webcasts, Podcasts, Outside Writing, and Conferences

Favorite Securosis Posts

  • Adrian Lane: Friday Summary: Decisions, Decisions. Rich is being modest here – he created a couple really cool tools while learning Chef and Puppet. Even some of the professional coders in the audience at BH were impressed. Drop him a line and give him some feedback if you can. There will be loads of work in this area over the coming months – this is how we will manage cloud security.
  • David Mortman: Dealing with Database Denial of Service.

Other Securosis Posts

Favorite Outside Posts

  • David Mortman: Busting the Biometric Myth. A vendor quoted some of it in a presentation.
  • Adrian Lane: Reducing Security. This made me laugh because it’s funny. Then I said to myself, “Oh crap, this is actually useful”. RSnake has done a nice job of a creating a simple checklist for web app security. Then I read the intro of the post and apparently he followed the same realization curve I did.

Research Reports and Presentations

Top News and Posts

Blog Comment of the Week

This week’s best comment goes to Ed, in response to Firewall Management Essentials: Introduction. And no offense Ed, but with that intro, we almost deleted your comment as blog spam. Links are one thing, but telling Mike his post is brilliant almost means a spambot.

Thank you so much for your brilliant article.

I just wrote a tutorial on how to set up Arno IPTABLES firewall. May be it may help someone to setup his own firewall based on IPTABLES. This is the best firewall. In my tutorial you can find some examples for a mail server and for a Proxy server using SNAT and port forwarding. The location of my tutorial is here: http://cosmolinux.no-ip.org/raconetlinux2/arno_iptables_firewall.html

I wish it is useful to someone.