Disaster recovery: Educating the market

Statistics reveal that two out of five businesses that experience a disaster, technological or otherwise, go out of business within five years. This fact was revealed at VERITAS Software`s recent Disaster Recovery seminars held in Cape Town and Johannesburg late last month.

The events, part of VERITAS` global campaign to raise awareness of how important disaster recovery (DR) preparedness is to business survival, also involved Compaq, Dell, Microsoft and Sun Microsystems. The purpose, at least in the local context, was to demonstrate how to create, implement and maintain an effective DR strategy and thereby ensure that, in the event of something disastrous happening, all critical data can be recovered rapidly.

The discussions centred on three primary elements of business continuity: disaster recovery, business recovery and contingency planning. Sheldon Hand, sales engineer at VERITAS SA, believes the key to identifying what precautionary steps should be taken lies in assessing the range of possible threats to the two most important assets of any organisation - people and information.

"These threats fall into three broad categories: local issues such as system failure as a result of hardware of software faults; logical problems including software bugs, viruses, corruption of data or accidental deletions; and site-related issues caused by natural disasters, terrorism, sabotage or even something as simple as a power outage," he says.

However, Hand stresses that the most often over-looked issue pertaining to DR is the cost of downtime to the business. "Meta Group recently researched the financial and retail industries in order to quantify these costs and the results were staggering. For example, a system outage of a mere 3.5 hours was estimated to cost retailers almost $200 000 (well over R2-million at the current exchange rate) while banks could expect to lose as much as $58-million from a system failure causing so short an outage.

"It`s an exponential curve from there, with downtime of 182 hours (8 days) costing banks as much as $3-billion and retailers almost $10-million. Beyond that, Meta says the costs were incalculable," he adds.

Neal Watkins and Jason Phippen, VERITAS Software DR specialists imported from the UK for the event, covered this in their presentation and elaborated on some solutions available to help prevent companies from going through such pain. Trenton Potgieter, senior system consultant at Dell SA, explored data availability at the site, application, data and platform levels, highlighting ways in which cater to these demands.

His presentation took the view that, in today`s global economy, computing systems must be available 24 hours a day, 365 days a year. Ensuring access to data requires eliminating risks that might cause downtime. Availability exists across many levels with each layer in the availability continuum providing additional levels of fault tolerance and/or recovery.

"By using the availability continuum as a framework, Dell has developed solutions using a broad spectrum of industry-standard components targeted at providing entry level, mind-range and high-end availability."

The Compaq presentation dealt with the issue of Disaster Tolerant solutions, which is says are reliant not only on robust between site links and effective data mirroring software but also on the stability and non-disruptive growth capacity of both the target and initiator sites. It explored the importance of the architectural design of a SAN as well as various technical and business factors which should be considered when completing the design and its implications for uptime, quick restore and ultimately disaster tolerance. Speaking on behalf of Sun Microsystems, Jorgen Nielsen, director, business continuity solutions at MGX, examined what is required to take enterprise wide recovery planning out of the data centre and establish business driven, holistic recovery strategies. He outlined a proven methodology that offered an understanding of the steps that need to be taken to ensure the continuation of critical business activities under any circumstances.

"Every business is different and financial impact of a system outage varies according to the reliance of the company on that system," says Hand. "As such, there is no `one-size-fits-all` DR strategy. Price too, is an important factor and this should be weighed against the possible cost of not having data and/or systems available for an extended period," he notes. Accordingly, the Disaster Recovery seminars covered technologies from simple tape backup to taking periodic disk snapshots, replication and mirroring, volume management and even clustering to provide redundancy at a system level. Application fail-over and SAN (storage area network) technology were also discussed.

"If there is one message we tried to get across, it was that there is no substitute for planning. Through assessment of the risks involved, any company can protect itself from the effects of unexpected systems downtime," concludes Hand.