Size an N+1 or N+2 cluster: units required, spare capacity, per-unit load after a failure and the cost of redundancy.
N is the number of units the load actually demands, and the redundancy level adds units on top so that failures do not reduce served capacity. The post-failure figure is the check that matters: if it exceeds 100%, the cluster is redundant on paper but overloads the survivors. N+1 that pushes the remaining units to 115% of rated capacity is not redundancy, it is a delayed cascading failure.
Redundancy Calculator (N+1, N+2)
N = ceil(load ÷ unit capacity); deployed = N + spares (or 2N for full duplication); post-failure load per unit = load ÷ (deployed − 1).
N = ceil(load ÷ unit capacity); deployed = N + spares (or 2N for full duplication); post-failure load per unit = load ÷ (deployed − 1). N is the number of units the load actually demands, and the redundancy level adds units on top so that failures do not reduce served capacity. The post-failure figure is the check that matters: if it exceeds 100%, the cluster is redundant on paper but overloads the survivors.
N+1 that pushes the remaining units to 115% of rated capacity is not redundancy, it is a delayed cascading failure.
This calculator takes 4 inputs: Total load to carry, Capacity per unit, Redundancy level, Cost per unit. The pre-filled defaults are a realistic starting point — replace them with figures from your own environment for a result you can act on.