Redundant server reliability (series + parallel + k-of-n)
A computed reliability block diagram for a server — a single power supply in series with a redundant fan pair and a 2-of-3 disk array. The engine reduces the success logic to one system-reliability figure and flags the power supply as the single point of failure.
For the reliability engineer sizing redundancy
What this shows
The three building blocks of reliability success logic in one diagram. The single power supply sits in series — everything depends on it. The two fans are in parallel, so the cooling subsystem survives either one failing. The three disks form a 2-of-3 array that tolerates one loss.
The number is computed, not drawn. The engine reduces each group — fans to 1 − 0.05² = 0.9975, the disk array to 3·0.97²·0.03 + 0.97³ = 0.99735, then multiplies the series chain — to a system reliability of ≈ 0.9849. It also derives each block's Birnbaum importance and flags the power supply as the single point of failure (red border): with no redundancy in series, its failure alone takes down the whole system, and it carries the highest importance — the first place to add redundancy.