Don't Ignore the HttpRequestValidationException

> Doing so could be... potentially dangerous
Cover Image for Don't Ignore the HttpRequestValidationException

Overview

No Sitecore implementation would be complete without an onslaught of the log error:

System.Web.HttpRequestValidationException: A potentially dangerous ______ value was detected

Where the blank might be Request.Cookies, Request.Form, Request.Path, Request.QueryString, or Request.Url.

This class of errors is extremely common and can be caused by a variety of factors, all of which stem from the fact that an incoming request contains what .NET perceives to be potentially dangerous. It is a security feature of .NET to prevent cross-site scripting attacks and other security issues.

While many devs consider this class of errors to be a nuisance, I argue that they deserve more attention than they are given.

Error Variations

A potentially dangerous Request.Cookies value was detected from the client

A potentially dangerous Request.Form value was detected

A potentially dangerous Request.Path value was detected

A potentially dangerous Request.QueryString value was detected

A potentially dangerous Request.RawUrl value was detected

A potentially dangerous Request.Url value was detected

How to Reproduce the Issue

For whichever class of this error you are trying to reproduce, include values that contain characters such as < > -- those which are commonly used to exploit XSS vulnerabilities. While you're at it, you can also test a variety of other characters such as:

< > " ' & ; ( ) [ ] { } ^ @ _ - ~ = ' : \/ ; ! $ # & \\ , + < .

Dangerous Values vs. Non-Dangerous Values

The phrase 'potentially dangerous' implies that the inputs may or may not actually be dangerous. This is a critical distinction that can be forgotten because, almost always, the inputs are dangerous because they are triggered by malicious bots doing script-kiddy tier reconnaissance. But, there are also cases where the inputs are not dangerous.

When performing analysis, ask these key questions:

  1. Is the value actually dangerous?
  2. Who triggered the error? Was it a bot or a human?
  3. Can WAF rules be adjusted to prevent this error?
  4. What was the action that triggered the error (simple GET request, form submission, etc)?

The Most Concerning Variation: Form Values

Let's focus on this variation:

A potentially dangerous Request.Form value was detected

If you inspect the path by which this error is triggered, you may find that it corresponds with a custom form or a Sitecore form (via the path /formbuilder).

If you are seeing this error on a form path, it means that your site likely returned 500 error response and that the form submission was not processed.

What if the form submission was a contact form? A lead generation form? A newsletter signup form?

Perhaps the most interesting example is a form with password field. Presumably, users should be able to enter almost any character in a password field (including < and >). If you are seeing this error get triggered by a form value that looks like it could be a password, then one or more of the following may be true: someone can't log in, set their password, reset their password, or create an account.

Conclusion

False positives of non-dangerous values can have brutal consequences. It can result in a loss of data, a loss of customer trust, and a loss of revenue.

Adjust your WAF rules to reduce the noise from the malicious bots. Be thoughtful when composing your WAF rules so as to not block non-dangerous values. When the errors come in after those adjustments, take them seriously. They may be a sign of a larger issue.

Stay dangerous,

MG

I chose these words

More Posts

Cover Image for JSS: Reducing Bloat in Multilist Field Serialization

JSS: Reducing Bloat in Multilist Field Serialization

> Because: performance, security, and error-avoidance

Cover Image for One Key Trick for Building Bulletproof Components That Work in Pages Editor

One Key Trick for Building Bulletproof Components That Work in Pages Editor

> Developers face the same key challenge when building for Pages as they did for the Experience Editor

Cover Image for Considerations for Hosting Mail Signature Images on Vercel

Considerations for Hosting Mail Signature Images on Vercel

> Outlook is a Cache-Control disrepectoor and that's a problem

Cover Image for NextJS: Unable to Verify the First Certificate

NextJS: Unable to Verify the First Certificate

> UNABLE_TO_VERIFY_LEAF_SIGNATURE