| Abstract | The Domain Name System (DNS) provides domain-to-address lookup-services used by almost all internet applications. Because of this ubiquitous use of the DNS, attacks against the DNS have become more and more critical. However, in the past, studies of DNS security have been mostly conducted against individual protocols and applications. In this thesis, we perform the first comprehensive evaluation of DNS-based attacks against a wide range of internet applications, ranging from time-synchronisation via NTP over internet resource management to security mechanisms. We show how to attack those applications by exploiting various weaknesses in the DNS. These attacks are based on both, already known weaknesses which are adapted to new attacks, as well as previously unknown attack vectors which have been found during the course of this thesis. We evaluate our attacks and provide the first taxonomy of DNS applications, to show how adversaries can systematically develop attacks exploiting the DNS. We analyze the attack surface created by our attacks in the internet and find that a significant number of applications and systems can be attacked. We work together with the developers of the vulnerable applications to develop patches and general countermeasures which can be applied by various parties to block our attacks. We also provide conceptual insights into the root causes allowing our attacks to help with the development of new applications and standards.
The findings of this thesis are published in in 4 full-paper publications and 2 posters at international academic conferences. Additionally, we disclose our finding to developers which has lead to the registration of 8 Common Vulnerabilities and Exposures identifiers (CVE IDs) and patches in 10 software implementations. To raise awareness, we also presented our findings at several community meetings and via invited articles. |
|---|
