It is easy to get carried away by the hype surrounding disaster
recovery but less hysteria and more sensible precautions, such as
backing up up data, are all that is needed.
Disaster recovery would be taken more seriously if it was called
irritation recovery. When it comes to disaster recovery the fact is
that few people outside of the military have changed their ways
since the terrorist attach in the US in September 2001.
The reason for this is simple. Disaster recovery is usually sold on
the false premise that every IT system is as vital as every other
system. But common sense tells us that not all systems rank in
importance with air traffic control, for example.
Disaster recovery even has the wrong name, and this being IT, it is
immediately given a second name: business continuity which, to
confuse matters further, we are told has as much to do with
security as avoiding disaster.
Users are told that if their systems crash, they will lose a
million pounds per moment and that in the Internet age every second
of downtime is costing you - big time. In extreme cases this may be
the case, but planning every time for the extreme worst-case
scenario will make for some very expensive plans. Business-critical
applications need to be highly protected and time-to-recovery
documented, but not all applications are business-critical.
The reality is that if a phone is busy it is only annoying - unless
you are dialling 999. How many of us take or make life-or-death
phone calls on a daily basis? Usually it is for something much more
mundane. And if the line is busy, you will call back. The same
applies to transacting - if you can not get through you will try
again.
Human nature is based on habit. If your favourite retail or
insurance company's Web site goes down, it is most likely that you
will wait for five minutes and try again instead of flouncing off
to a competitor. People are driven by habit and familiarity backed
by a (usually misguided) sense of loyalty.
Having 100% guaranteed uptime is a nice idea. The fact is that even
at the very top end of the market there are businesses for whom
mainframe technology is not robust enough.
Remember Tandem Computers, once an independent company, then part
of Compaq and now part of Hewlett-Packard? There was a time when
Tandem made a nice living selling fault-tolerant computers - it
made so nice a living that each Friday afternoon staff at its
Cupertino headquarters downed tools and drank beer.
Tandem sold fault-tolerant mainframes. Big tin - effectively two
mainframe computers bolted back to back. If one stopped working,
the other kept going. It worked. They did not sell very many (they
did not need to) but it was perfect technology for those companies
that needed it.
Unfortunately disaster recovery is being pushed in the same way but
how many businesses really need a couple of mainframes in the
basement?
Should you have a recovery plan? Of course. Should you back-up and
protect data? Naturally. And should someone have responsibility for
it? By all means. Remember if something goes wrong and more than
one person is responsible, no-one takes the blame. Well, you can do
your best, but failure will always be an orphan.
Since last September the spotlight has been put on disaster
recovery, but for most of us outside the Ministry of Defence or the
nuclear power industry disaster recovery plans should be mundane,
and routine.
Disaster is always much more exciting than the recovery and
suppliers try to sell on the potential for disaster.
As a consequence you always read about disasters but rarely about
recovery, which is dull but worthy. It is not about saving the day,
it is simply about having data backed-up and readily accessible
through well planned business processes.
If you want to avoid disaster, lock up at night, backup at night
and do not allow your entire board of directors to travel on the
same helicopter.
Ian French is president of Bell Microproducts Europe