Investment in disaster recovery and business continuity services has traditionally been a grudge purchase, says Brian Timperley of Turrito Networks — but with advances in virtualisation and cloud-based application and data delivery, it’s now possible to get better results at a much lower cost.
“Traditional business continuity services in South Africa are crazy expensive,” says Timperley. “There are few companies that can make a viable business case for maintaining a second set of replicated servers at a second site. As a result, we’ve noticed many companies deciding just to live with the risk.”
Even those who do make the investment face difficult choices when disaster actually strikes. “If you’ve gone to the additional trouble and expense of testing your disaster recovery, you can at least be confident it will work;” says Timperley; “but once you’ve failed over, getting the original site up and running again is not trivial. Deciding exactly when to push the big red button is not for the faint-hearted.”
What companies should be aiming for instead, argues Timperley, is disaster resilience — a failover so seamless that users don’t even notice it.
“We had a perfect example of how disaster resilience can work just recently, while doing a proof of concept for our Cloudware application delivery system with a large client in the metal industry,” says Timperley. “A power surge damaged the UPSs in their primary data centre, and prevented their generators from kicking in: The entire data centre switched off.”
The client did have a second site where their data was safely replicated, “but no way of remotely accessing it, and no way of allowing their admin staff to continue working without interruption,” says Timperley. “One of their senior IT staff switched the Cloudware licences from the primary to the secondary site, which took all of 40 minutes to bring their business back to normal. Users were able to access their applications, and continue to work, while IT sorted out the primary data centre.”
“Companies should no longer be asking themselves whether they can successfully fail over to a secondary site in the event of a disaster,” says Timperley. “The new question should be: How can I build resilient systems that reliably deliver applications and data to my users even when disaster strikes? Even if your offices burn down, you should be up and running again as soon as your employees can find a working Wifi connection or 3G signal.”
Timperley notes that Gartner, at its recent Africa Symposium in Cape Town, predicted that one of the top ten strategic technology trends for 2015 would be “web-scale IT” — building and delivering like Amazon or Google even within the enterprise. “The global web giants are built for resilience. That’s increasingly the way every business is going to build its IT.”