システムや機器の信頼性に関する記述のうち, 適切なものはどれか。
平成27年度 春期 ITパスポート試験 問64
This question's explanation and your answers so far.
問 64
Incorrectア
The term does not match the description. Keeping damage to a minimum when a failure occurs and controlling the system into a safe state is called fail-safe (for example, a traffic light turning red in every direction when it breaks down, or a heater putting out its flame when it tips over). Fool-proof refers to a design that keeps a user's incorrect operation from leading to danger or failure (for example, a microwave oven that will not heat unless the door is closed).
Incorrectイ
The term does not match the description. Using high-quality, highly reliable parts and components so as to lower the probability that a failure occurs at all (preventing failures before they happen) is called fault avoidance, not fail-safe. Fail-safe is the idea of steering the system toward the safe side after a failure has occurred.
- ウ
正解 The correct answer is ウ. Fault tolerance means designing a system so that, even if a failure or fault occurs in part of it, the system as a whole can keep processing without stopping. A typical example is a redundant configuration in which servers, disks, power supplies and so on are duplicated or multiplexed, so that if one of them breaks the system switches over to the standby unit and keeps running. Choice ウ, “enabling a system to continue processing when a fault occurs in it due to a failure or the like,” is exactly this definition, so it is an appropriate description. It also helps to organize fault avoidance — trying to keep failures from happening at all — and fault tolerance — continuing to process on the assumption that failures do happen — as a matched pair of ideas.
Incorrectエ
The term does not match the description. Taking measures at the design stage so that people do not misoperate a system, or so that a mistake does not lead to a failure or fault, is called fool-proof. Fail-soft means cutting off the failed part when a fault occurs and continuing the minimum necessary operation even at reduced performance or function (degraded operation, fallback); it is not a term about guarding against human operating errors.
- Tell the four key reliability-design terms apart: fault tolerance (keep processing even when a fault occurs), fail-safe (control the system to the safe side when it fails), fail-soft (keep running even at reduced function, that is, degraded operation), and fool-proof (prevent accidents on the assumption that people will operate things incorrectly).
- Fault avoidance (keeping failures from occurring in the first place) and fault tolerance (protecting operation on the assumption that failures will occur) are a matched pair of concepts, and adopting highly reliable parts belongs to the former.