Byzantine Failure and Resilience
Failures in an algorithm or protocol can be categorized into three main types:
- A failure to take another execution step in the algorithm: This is usually referred to as a "fail stop" fault.
- A random failure to execute correctly: This is called a "random fault" or "random Byzantine" fault.
- An arbitrary failure where the algorithm fails to execute the steps correctly (usually in a clever way by some adversary to make the whole algorithm fail) which also encompasses the previous two types of faults; this is called a "Byzantine fault".
A Byzantine resilient or Byzantine fault tolerant protocol or algorithm is an algorithm that is robust to all the kinds of failures mentioned above. For example, given a space shuttle with multiple redundant processors and some of the processors give incorrect data, which processors or sets of processors should be believed? The solution can be formulated as a Byzantine fault tolerant protocol.
Read more about this topic: Quantum Byzantine Agreement
Famous quotes containing the words failure and/or resilience:
“Any effort in philosophy to make the obscure obvious is likely to be unappealing, for the penalty of failure is confusion while the reward of success is banality. An answer, once found, is dull; and the only remaining interest lies in a further effort to render equally dull what is still obscure enough to be intriguing.”
—Nelson Goodman (b. 1906)
“Toddlers who don’t learn gradually about disappointment lose their resilience through lack of practice in give-and-take with other people’s needs. They can become self-centered, demanding, and difficult to like or to be with.”
—Alicia F. Lieberman (20th century)