DCSW key terms

The ability to service or repair one component of a system while the rest of the system continues operating at full design capacity, without interrupting the critical load.

A design property where the system keeps supporting the critical load despite the failure of any single component, achieved through redundancy and automatic transfer rather than manual intervention.

The arrangement of spare capacity in a system. N+1 provides one extra module beyond the minimum needed; 2N provides two complete independent paths, each capable of carrying the full load.

A four-level ranking of data center site infrastructure by availability and fault tolerance, from Tier I (basic, no redundancy) to Tier IV (fault tolerant and concurrently maintainable).

A widely referenced framework that certifies data center design documents and constructed facilities against defined topology and performance criteria for availability, using four tiers.

A North American standard for data center telecommunications infrastructure, covering site location, architectural, electrical, mechanical and cabling requirements, with its own four-tier rating scheme.

The portion of electrical demand that must never be interrupted, typically IT equipment and its supporting cooling, which drives the sizing and redundancy of all upstream power and mechanical systems.

A device that supplies conditioned power from batteries or another stored source during a utility outage, bridging the gap until generators start and stabilize.

A switch that senses loss of the normal utility source and automatically transfers the load to an alternate source such as a generator, then back again when utility power returns.

A solid-state switch that transfers a load between two live power sources in a few milliseconds, far faster than a mechanical breaker, protecting IT equipment from voltage sags.

The synchronized operation of two or more generators onto a common bus so they share load, allowing capacity to scale and one unit to be serviced while others carry the demand.

The practice of applying artificial electrical load to generators, UPS systems or cooling plants to verify they can carry rated capacity under controlled, measurable conditions.

The ratio of total facility power consumption to the power delivered to IT equipment. A value of 1.0 is theoretically perfect; lower numbers indicate more efficient infrastructure.

A layout strategy that physically separates server intake and exhaust air streams using panels or doors, preventing hot and cold air from mixing and improving cooling efficiency.

The deliberate control of air paths in a data hall through blanking panels, raised floor tiles, seals and containment so that cold supply air reaches equipment intakes without bypass or recirculation.

Computer modeling of air temperature, velocity and pressure across a data hall, used to predict hot spots and test cooling layouts before equipment is installed.

A refrigeration machine that removes heat from a cooling water loop and rejects it outdoors, forming the core of a chilled water cooling plant serving large data centers.

A heat-rejection device that evaporates water to discharge condenser heat to the atmosphere, typically paired with chillers or water-cooled equipment in large facilities.

A mode of operation that uses cool outdoor air or cold outdoor water instead of mechanical refrigeration to cool the facility, cutting compressor energy use in cool climates.

A elevated floor structure whose void beneath serves as a pathway for cold supply air, power cabling or both, distributing air through perforated tiles into the data hall.

A graph relating air temperature, humidity, moisture content and enthalpy, used to analyze cooling coil performance and confirm supply air stays within equipment allowable envelopes.

The range of temperature and humidity conditions within which IT equipment manufacturers permit operation, defining the boundary that cooling systems must keep the data hall inside.

The temperature at which air becomes saturated and water vapor condenses; keeping supply air above the dew point prevents condensation on equipment and server internals.

A standardized, hierarchical cabling system using defined media, connectors and pathways that supports voice, data and control signals independent of the equipment attached to it.

A patching point where cabling from different parts of the network is terminated and interconnected, such as the main distribution area or a horizontal distribution area.

A controlled space where external carriers terminate their fiber and interconnect with the facility's internal network, serving as the entry point for telecommunications services.

High-bandwidth optical cabling that links major distribution areas and carries aggregated traffic between floors, halls and carrier entry points, offering immunity to electromagnetic interference.

A fire suppression system using electrically non-conductive gaseous agents that extinguish fire without leaving residue, protecting IT equipment from water or chemical damage.

An aspirating smoke detection system that continuously draws air samples through pipes to a sensitive laser detector, identifying incipient fire before visible smoke appears.

A detection logic requiring two independent detectors to alarm before suppression discharges, reducing the chance of a false activation that would disrupt operations.

Fire-rated walls, floors, doors and penetration seals that contain fire and smoke within a compartment for a defined period without active suppression.

A small controlled vestibule with two interlocking doors, where one must close before the other opens, preventing tailgating and controlling access into a secure area.

The security breach in which an unauthorized person follows an authorized individual through a controlled door without presenting credentials, a key reason for vestibules and turnstiles.

The integrated hardware and software that authenticates identities via cards, biometrics or PINs and electronically controls door locks, logging every entry and exit event.

A structured end-to-end test that proves power, cooling, fire and monitoring systems work together under both normal and failure scenarios before the facility is accepted.

The documented process of verifying that installed systems meet design intent and owner requirements, from factory testing through functional performance tests and handover.

A written description of how systems respond to conditions and failures, such as generator start, transfer and load sequencing, used to program controls and verify correct behavior.

The maximum acceptable duration to restore a business process or IT service after disruption, guiding the design of failover capability and staffing response.

The maximum acceptable amount of data loss measured in time, determining how frequently backups or replication must occur between sites.

A structured study identifying critical business functions, their dependencies and the financial and operational consequences of downtime, used to set recovery objectives.