denial of service

An attack that makes a service unavailable to its real users by overloading it or making it crash.

An outage stops every user, and algorithmic variants need very little attacker bandwidth.

In more detail

Some attacks flood a service with traffic; algorithmic attacks such as hash flooding instead send small inputs that trigger worst-case behaviour. Defences include input limits, rate limiting and algorithms whose worst case is acceptable. When the attack comes from many machines at once it is called distributed denial of service.

Examples

The course shows that an attacker can spend your resources without reading any data, for example with unlimited login guesses or a request for a million rows, which is why rate limits and size bounds belong at the API boundary.
Amplified DDoS abuses open services such as DNS or time servers that send large replies to small spoofed requests, aiming the replies at the victim.

Related terms

Taught in

References

Certification courses