Week of War on WAF's: Day 2 -- A look at the past
Web application experts have been asking WAF vendors the same questions for years with no resolution. It's not about religion for many security professionals -- it's about having a product that works as advertised.
My frustration is not unique. I am not the first person to clamor on about web application firewalls. Jeff Williams pointed me to a post that Mark Curphey made in 2004. Today, Curphey appears to have a change of heart -- his latest blog post provides a link to URLScan, which some claim is like mod-security for Microsoft's Internet Information Server (IIS). Microsoft released URLScan Beta 3.0 in order to curtail the massive problem of over two million Classic ASP web applications that have become infected due to the SQL injection attacks.
Here is the post where the frustration of WAF and their vendors first began:
blog comments powered by Disqus
- -----Original Message-----
- From: The OWASP Project [mailto:[EMAIL PROTECTED] Sent: Tuesday, 16 November 2004 2:34 PM To: [EMAIL PROTECTED] Subject: An Open Letter (and Challenge) to the Application Security Consortium
An Open Letter (and Challenge) to the Application Security Consortium
Since its inception in late 2000 the Open Web Application Security Project (OWASP) has provided free and open tools and documentation to educate people about the increasing threat of insecure web applications and web services. As a not-for-profit charitable foundation, one of our community responsibilities is to ensure that fair and balanced information is available to companies and consumers.
Our work has become recommended reading by the Federal Trade Commission, VISA, the Defense Information Systems Agency and many other commercial and government entities.
The newly unveiled Application Security Consortium recently announced a "Web Application Security Challenge" to other vendors at the Computer Security Institute (CSI) show in Washington, D.C. This group of security product vendors proposes to create a new minimum criteria and then rate their own products against it.
The OWASP community is deeply concerned that this criteria will mislead consumers and result in a false sense of security. In the interest of fairness, we believe the Application Security Consortium should disclose what security issues their products do not address.
As a group with a wide range of international members from leading financial services organizations, pharmaceutical companies, manufacturing companies, services providers, and technology vendors, we are constantly reminded about the diverse range of vulnerabilities that are present in web applications and web services. The very small selection of vulnerabilities you are proposing to become a testing criteria are far from representative of what our members see in the real world and therefore do not represent a fair or suitable test criteria.
In fact, it seems quite a coincidence that the issues you have chosen seem to closely mirror the issues that your technology category is typically able to detect, while ignoring very common vulnerabilities that cause serious problems for companies.
Robert Graham, Chief Scientist at Internet Security Systems, recently commented on application firewalls in an interview for CNET news. When asked the question "How important do you think application firewalls will become in the future?" his answer was "Not very."
"Let me give you an example of something that happened with me. Not long ago, I ordered a plasma screen online, which was to be shipped by a local company in Atlanta. And the company gave me a six-digit shipping number. Accidentally, I typed in an incremental of my shipping number (on the online tracking Web site). Now, a six-digit number is a small number, so of course I got someone else's user account information. And the reason that happened was due to the way they've set up their user IDs, by incrementing from a six-digit number. So here's the irony: Their system may be so cryptographically secure that (the) chances of an encrypted shipping number being cracked is lower than a meteor hitting the earth and wiping out civilization. Still, I could get at the next ID easily. There is no application firewall that can solve this problem.
With applications that people are running on the Web, no amount of additive things can cure fundamental problems that are already there in the first place."
This story echoes some of the fundamental beliefs and wisdom shared by the collective members of OWASP. Our experience shows that the problems we face with insecure software cannot be fixed with technology alone. Building secure software requires deep changes in our development culture, including people, processes, and technology.
We challenge the members of the Application Security Consortium to accept a fair evaluation of their products. WASP will work with its members (your customers) to create an open set of criteria that is representative of the web application and web services issues found in the real world. OWASP will then build a web application that contains each of these issues. The criteria and web application will be submitted to an independent testing company to evaluate your products.
You can submit your products to be tested against the criteria (without having prior access to the code) on the basis that the results are able to be published freely and will unabridged.
We believe that this kind of marketing stunt is irresponsible and severely distracts awareness from the real issues surrounding web application and web services security. Corporations need to understand that they must build better software and not seek an elusive silver bullet.
We urge the Consortium not to go forward with their criteria, but to take OWASP up on our offer to produce a meaningful standard and test environment that are open and free for all.
- Contact: [EMAIL PROTECTED]
- Website: www.owasp.org