Demystifying Common Data Center Terms and Acronyms

Like any technical field, the data center industry has unique terminology. Many terms are shortened to acronyms that can confuse even industry pros. New terms and acronyms are constantly being added. Older ones can catch younger IT professionals off guard.

We’ve covered common data center performance metrics and sustainability terms and concepts in separate blogs. Here, we’ll explore frequently used industry jargon.

Data Center Infrastructure

ATS (Automatic Transfer Switch). The ATS automatically detects a power outage and transfers the load to the backup power source. It then transfers the load back to the primary power source when utility power is restored.

BMS (Building Management System). A BMS monitors and controls mechanical, electrical, and electromechanical services such as HVAC, power, access controls, and lighting.

CRAC (Computer Room Air Conditioning). CRAC units are the traditional HVAC cooling systems for the data center. They remove hot air from the room, chill it, and return it to the room.

CRAH (Computer Room Air Handling). CRAH units serve the same purpose as CRAC units. However, they use chilled water and valves instead of refrigerant and compressors and are typically reserved for large deployments because they require significant infrastructure changes. They’re more expensive upfront but have long-term efficiency benefits and savings. 

DCIM (Data Center Infrastructure Management). DCIM tools aid in the monitoring and management of data center infrastructure and provide metrics on energy usage, utilization, etc. 

EPO (Emergency Power Off). The EPO system comprises one or more wall-mounted buttons that enable rapid shutdown of power to one or more pieces of equipment or the entire data center.

HVAC (Heating, Ventilation, and Air Conditioning). HVAC refers to all the systems and technologies used to control the temperature and humidity of the air inside the data center.

PDU (Power Distribution Unit). A PDU is a device with multiple outlets that takes power from a single input source and supplies it to multiple devices. Rack mount PDUs are essentially large power strips that may have intelligent features such as monitoring and metering.

UPS (Uninterruptible Power Supply). If the primary power source fails, a UPS supplies backup power to a computer system for a short time to enable controlled shutdown. Rechargeable batteries are the most common source of power storage. Rack mount UPSs are common in small server rooms and network closets, while large enterprise data centers commonly use more sophisticated systems.

Colocation. Also known as a “colo” or multi-tenant data center (MTDC), a colocation data center is a facility where businesses can rent infrastructure space to host their computer environment without the need for capital expenditure (CapEx) in infrastructure cost. 

Meet Me Room (MMR). A room within a colocation data center where telecommunications companies can physically connect, exchange data, and peer into each other's multiple networks.

Main Distribution Frame (MDF). An area to connect both private and public service lines that come into the data center with customers' networking devices within the data center. More on MMRs, MDFs, and more in our data center white space vs. gray space blog.

Carrier Neutral Facilities. Colocation data centers that are non-affiliated with any one particular telecommunication service provider. They provide customers with multiple network services.

On-Premises (On-Prem). IT infrastructure that is hosted locally on-site outside of a colocation data center.

N+1, N+2, 2N, 2N+1 (Redundancy Levels). Redundancy levels are described in reference to the baseline “N,” which refers to the minimum number of independent resources a system needs to operate. N+1 includes one independent backup system. N+2 has two backup systems, while 2N doubles the resources available to the system. 

Networking

CDN (Content Delivery Network). A CDN is a geographically distributed network of servers that cache content close to users. It is designed to deliver content quickly, cheaply, securely, and reliably.

IPv4/IPv6 (Internet Protocol version 4/6). IPv4 is the fourth version of the Internet Protocol and the primary routing mechanism for Internet traffic. It uses 32-bit addresses. IPv6 is the latest version. It uses 128-bit addresses, enabling a far larger address space.

LAN (Local Area Network). A LAN is a network that connects computer systems and other devices within a single facility.

SDN (Software-Defined Networking). An SDN uncouples the network control and data planes, enabling dynamic, software-driven network configuration.

VLAN (Virtual LAN). A VLAN is a logical network that subdivides a LAN and isolates the traffic in each subdivision.

VPN (Virtual Private Network). A VPN creates a secure “private tunnel” in a public network by encrypting the data and communications traveling over the connection.

WAN (Wide Area Network). A WAN is a network that connects computer systems and other devices over a large geographic distance, generally using leased telecom circuits.

Compute, Storage, and Virtualization

DAS (Direct Attached Storage). DAS refers to a storage device that is directly connected to a server.

HCI (Hyperconverged Infrastructure). HCI systems virtualize compute, storage, and networking on a single commodity server. A management plane enables easy administration of resources from one interface.

HPC (High-Performance Computing). HPC uses clusters of computers or computers with clusters of processors working together to perform complex calculations at high speed.

NAS (Network Attached Storage). A NAS is a network-connected file server that provides data access to multiple systems.

RAID (Redundant Array of Independent Disks). RAID combines multiple physical disk drives in a single logical unit, enabling data replication to improve reliability performance.

SSD (Solid State Drive). Also known as flash storage, an SSD uses nonvolatile flash memory instead of a spinning disk for storage.

VDI (Virtual Desktop Infrastructure). With VDI, the desktop environment and applications run as software in the data center and are delivered over the network to the client device.

VM (Virtual Machine). A VM is a software emulation that provides the basic functionality of a physical computer system.

General

BCP (Business Continuity Plan). A BCP documents an organization’s strategy for remaining operational in a major disruption.

DRP (Disaster Recovery Plan). A DRP is a structured approach to recovering from a major disruption.

ISP (Internet Service Provider). An ISP provides the services required to access and use the Internet.

RPO (Recovery Point Objective). The RPO is the tolerable amount of data that can be lost in a system failure or disaster. It is measured as the time between data backups.

RTO (Recovery Time Objective). The RTO is the tolerable amount of time a system can be down after a failure or disaster. It is measured as the time required to recover data.

SLA (Service Level Agreement). An SLA defines the level of performance and reliability that a customer can expect from a provider.

Enconnex Is Your Source for Data Center Infrastructure

Enconnex offers an array of data center infrastructure solutions to meet a wide range of needs. Let our specialists help you decipher the alphabet soup and find the best solutions for your initiative. Get in touch today.