Disaster recovery site options

When planning, setting up and operating a disaster recovery (DR) plan, having alternate sites for IT operations is critical. In this tip, Ed Tittel discusses the specific features, functions and costs of hot, warm, cold and mobile sites.

When planning, setting up and operating a disaster recovery (DR) plan, having alternate sites for IT operations is critical. Disaster recovery site options include hot sites, warm sites, cold sites and mobile sites. Each has particular features, functions and costs.

Hot sites

A hot site is a full or partial duplicate for a primary IT operation, including complete computer systems and near-real-time backups for systems, applications and data. In its most expensive form, mirroring software is used to keep a hot backup site and a primary site synchronized.

A hot site is used when an organization can tolerate little or no downtime. Switchover typically takes no more than a few hours, and may occur more quickly than that. Because staff must travel from a primary to a hot site after a disaster, data and services may be available to users sooner than staff. Hot sites make sense for businesses where ongoing operations are critical, or where costs of downtime (calculated using standard risk assessment) meet or exceed costs for hot site acquisition, outfitting and maintenance. This includes government agencies (especially FEMA and the DoD), financial institutions and large e-commerce and trading operations (stock or commodities markets).

Warm sites

A warm site is best understood as a hot site minus data replication. That is, a warm site offers access to space, utilities and equipment, but requires current backups be installed, and systems and services brought online to become operational. A warm site may be a complete duplicate of an original site, but will typically provide only a subset of mission-critical equipment, services and data.

A warm site works for businesses or organizations that can tolerate one or two days of downtime, which represents the typical delay between when a primary site goes down and a recovery site comes up. Many major and medium-sized businesses opt for warm sites because costs are significantly lower than for hot sites. Although equipment and siting costs are similar, data synchronization and ongoing maintenance and monitoring costs don't apply to warm sites.

Cold sites

A cold site is best understood as a DR plan and arranges to enact that plan when necessary. A business or organization typically makes arrangements for access to a recovery site with adequate utilities and services, and to purchase necessary equipment to mount and restore essential IT operations. Aside from costs of planning and arranging for DR, and access to sufficient funds or lines of credit necessary to cover equipment acquisition and siting costs, cold sites don't require substantial up-front outlays.

Bringing a cold site up usually takes one to two weeks, assuming that a site, equipment and backups can be acquired and activated that quickly. This can take longer, particularly after a natural disaster (think about New Orleans businesses following Katrina) when many businesses must compete to meet similar needs. That's why cold sites make sense primarily for small-midsized businesses (SMBs) that can continue to function for some time without a fully operational IT infrastructure.

Mobile sites

A mobile site sits somewhere between a warm and a cold site, where a site operator makes portable structures equipped with computing equipment available to customers. The degree to which computing infrastructure is decided and funded in advance decides the relative "temperature" of a mobile site (more advance funding and preparation is warmer, less is colder).

Bringing a mobile site online depends on how quickly it can be delivered to its desired location, and how quickly backups are restored and made operational. Because mobile sites are limited in scope and scale, they make most sense for SMBs, albeit those with less tolerance for delay in resuming IT operations.

Cost factors

Costs may be understood in terms of purchase and monthly expenditures for a basic IT server at a recovery site, plus square footage for housing; cooling, power, utilities and high-speed Internet access (100 Mbps or better). Costs and capabilities scale as the number of units involved increase. Up-front acquisition and ongoing maintenance and operation costs make the biggest difference between hot, warm and cold disaster recovery sites (hot and warm sites must upgrade and replace recovery gear in synch with primary gear). Dollar entries compare relative costs (for every dollar spent on cold sites, plan to spend $2.50 on warm, and $10 on hot equivalents).

About this author: Ed Tittel writes regularly for numerous TechTarget Web sites on networking, IT security, and developer topics. His most recent books are "Windows 2008 Server for Dummies"(Wiley, 2008) and the "CISSP Study Guide, 4e" (Sybex, 2008). 

Dig Deeper on Disaster recovery facilities - operations