Defensive programming

This paper describes a toolkit to help improve the robustness of code against DoS attacks. We observe that when developing software, programmers primarily focus on functionality. Protecting code from attacks is often considered the responsibility of the OS, firewalls and intrusion detection systems. As a result, many DoS vulnerabilities are not discovered until the system is attacked and the damage is done. Instead of reacting to attacks after the fact, this paper argues that a better solution is to make software defensive by systematically injecting protection mechanisms into the code itself. Our toolkit provides an API that programmers use to annotate their code. At runtime, these annotations serve as both sensors and actuators: watching for resource abuse and taking the appropriate action should abuse be detected. This paper presents the design and implementation of the toolkit, as well as evaluation of its effectiveness with three widely-deployed network services.