Business continuity part 2: Moving quickly when disaster strikes

15 Jan 2004

There is something incongruously romantic about the idea of ‘hot sites’ for an organisation to recover quickly again after a disaster, with visions of wartime command bunkers and the camaraderie of emergencies. They are usually seen as the prerogative of very large organisations – or very rich activities such as financial dealing rooms – which can afford to subscribe to such luxuries in planning for business continuity.

In practice, they are usually prosaic and nondescript buildings in industrial estates, without much by way of identification because anonymity is one small part of the detailed attention they must pay to their own security. A disaster-recovery service, after all, has itself to be as disaster-proof as possible. Inside they are really just data centres with lots of spare office space.

Not all business disasters involve loss of premises. The classic Dublin tale is of the Fitzwilliam Square firm some years ago where the Fire Brigade successfully dealt with a fire in a top-floor flat. The three floors below the business hardly had a smell of smoke, but the water wrecked the network servers, telephone switches and a crucial PC with confidential data not on the server. With both IT system and phones gone in a case like that, your people have somewhere to sit but you’re out of business until replacement equipment is installed and running.

“Replacing equipment and restoring data and applications is the commonest solution for business continuity,” says Brian Hurley, managed services manager of Hewlett-Packard (HP) Ireland. “With a disaster-recovery contract everything can be moved speedily. Your service level agreement may deliver replacement servers and so on in four hours and the restoration of your data and applications will have been planned and rehearsed so you will know how long it is likely to take.” He suggests, however, that businesses may tend to underestimate the overall time such a process could take. With a total outage, eight hours would be good going and 24 hours or more quite likely before a business IT system was completely restored and operational, although he stresses that in the nature of disasters there is no ‘typical’ scenario.

All business continuity planning is based on assessing risk vs costs. The top-end IT solution is having your entire database and applications mirrored in real-time on an alternative site – your own or a third-party data centre. If the unthinkable happens, everything is ready to keep rolling provided your staff have the communications and devices. For automated applications, such as website hosting and interaction and supply chain links (an increasingly crucial element in continuity because of the potential knock-on impact on trading partners), major elements of the business may in fact continue seamlessly even as the fire rages or whatever. As people move to their pre-planned alternative locations there may be some breaks in service but most customers and contacts will not have occasion to notice. The standby office solution could be a hot site, designated emergency accommodation within the organisation or even, in these days of e-working, dispersal of senior staff to home offices.

In many respects the key point about the information and communications technology (ICT) elements of business continuity is that despite their total criticality to any business they are just one part of a business continuity plan. Yet it is fair to say that there is more accumulated knowledge, experience and packaged solutions on the ICT side because the disaster recovery hot site emerged as technology became mission-critical in the Eighties.

IBM has two data centres-cum-disaster recovery sites on the Dublin outskirts, but Tom Walsh, manager of the service, readily points out that a hot-site solution is the very top end of a range of measures that could be appropriate for different businesses.

“The days when people had to go where the systems were are long gone, so in these days of the web and distributed computing there are many possible solutions. Quite simply the only way to decide is through detailed assessment of the risks and the potential downside costs to the business that can then be weighed against the likely costs of the elements in an effective continuity plan. Technically, there really are few problems in implementing proven ICT solutions once the policies and priorities are set,” Walsh explains. There is a growing trend, he says, for ICT continuity plans to be a natural extension of managed services, now the standard approach to looking after servers, local area networks and wireless area networks in so many Irish organisations with minimal in-house expertise.

This is echoed by HP’s Brian Hurley, who points out that equipment simply is not the problem. “A small van can have a whole rack of servers, PCs and peripherals on a designated site in a matter of hours and a single technician can get them up and running equally quickly. The real technical secrets are in the design and knowledge of the systems and applications so that they can actually be restored from backup to full functionality. That sort of expertise is precisely what a good managed-service provider offers in the first place.”

By Leslie Faughnan