Ask the Expert

Best practices for managing DNS, knowing it's anything but trustworthy

What would you say are some best practices for managing DNS now that we know that it's anything but trustworthy? More specifically, where should DNS security live within the average enterprise's risk matrix?

    Requires Free Membership to View

    Login

If there is anything that Dan Kaminsky's research on how to break the DNS has shown, it's that organizations are pretty exposed. But security pros need to understand there are a few things that can be done to mitigate the risk. The first practice is to make sure the company's DNS server practices source-port randomization. This doesn't totally solve the Kaminsky DNS attack, but it makes it hard enough (and costly enough) to largely mitigate the risk.

That means the DNS server must be patched, or the company should upgrade to a more robust server infrastructure. One possibility is DNSSEC, which the U.S. federal government just deployed, but that is pretty complicated.

Another important thing is to make sure that all upstream ISPs have their act together. Even if the company's system is fine, dealing with any compromised name servers is disastrous.

In terms of where the responsibility of DNS security should reside, that depends on what kind of operational responsibilities the security team has. Many security teams these days are more influencers than implementers, which means they need to work with the organization's network team, which would actually deploy any remediation.

More information:

This was first published in July 2008

Join the conversationComment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.
    Back to top