問67
頻出度★★★
システムの一部に障害が発生した場合でも, 正常に処理を実行することができる施策として, 適切なものだけを全て挙げたものはどれか。 ① HDD をミラーリングで構成する。 ② システムの安定稼働後は, 保守や点検の頻度をできるだけ減らす。 ③ 冗長化していた複数のネットワーク回線を, より高速な 1 本の回線にまとめる。 ④ 無停電電源装置を設置するなどして電源を多重化する。
平成28年度 秋期 ITパスポート試験 問67
テクノロジ系システム構成要素 › システムの評価指標
This question's explanation and your answers so far.
解説
問 67
Incorrectア
(1) is appropriate, but (3) is not. Option (3) consolidates the previously redundant multiple network lines into a single higher-speed line; once only one line remains, a single point of failure appears, and if that line goes down, processing stops. Because it throws away redundancy, it is wrong as a fault-tolerance measure.
- イ
正解 The correct answer is “イ”（(1) and (4)）. This question asks you to choose a design that keeps processing running even when part of the system breaks, that is, a “fault tolerant” design achieved through redundancy and duplication. (1) Configuring the HDD with mirroring (RAID1) writes the same data to two HDDs, so even if one fails, processing can continue on the other. (4) Making the power supply redundant with an uninterruptible power supply (UPS) and similar devices keeps power flowing without interruption even during a blackout or power-system failure, so processing can continue. Both (1) and (4) are redundancy measures that “cover a partial failure with another part”, so “イ”, which lists both, is appropriate.
Incorrectウ
Both (2) and (3) are inappropriate. Option (2), reducing the frequency of maintenance and inspection, makes it easy to overlook signs of failure and actually lowers reliability. Option (3) consolidates the redundant lines into one, creating a single point of failure. Both run counter to the goal of keeping processing running when a failure occurs.
Incorrectエ
(4) is appropriate, but (2) is not. Option (2), “reduce the frequency of maintenance and inspection as much as possible after stable operation”, weakens early detection and prevention of failures and lowers availability, running opposite to the goal of keeping processing running when a failure occurs. Any combination that includes (2) is wrong.
ここがポイント
- Fault tolerant: the idea of making components redundant and duplicated so that processing can continue even when part of the system fails (e.g., HDD mirroring = RAID1, power-supply duplication with a UPS).
- A measure that gives up redundancy (consolidating multiple lines into one) creates a single point of failure, and cutting back on maintenance and inspection lowers reliability because signs of failure are missed; both undermine fault tolerance.
用語