N+1

N+1 is a redundancy class in which the system contains one additional component beyond the minimum (N) required to carry the design load. If a single component fails or is taken offline for maintenance, the remaining components carry the full load.

N+1 is the most common redundancy class in modern enterprise data centers. It is enough to survive any single component failure or planned outage of one component. It is not enough to survive simultaneous failures of two components, which is what 2N or 2N+1 topologies are designed for.

The interesting question is always: N+1 of what, and at what level? A facility can be N+1 at the UPS level, the chiller level, and the CRAH level, but only N at the utility feed or generator paralleling switchgear. Marketing-grade N+1 claims sometimes obscure these dependencies. Engineering due diligence catches them.

N+1 is the topology of choice for most colocation facilities because it balances reliability against capital cost. Hyperscale facilities frequently run 2N or block-redundant designs because the operational discipline behind N+1 is harder to maintain at scale.