eCommerce needs a business continuity system that functions so well customers don’t even notice disasters.
Data centre disasters are rare in the UK; fire precautions are generally very good and no-one builds data centres on flood plains without making sure they are well protected. But freak weather happens and power outages occur, and may become more common.
Many smaller organisations that operate 9 to 5 and close for the weekend can survive IT disasters by muddling through the day and fixing it overnight. eCommerce and financial are increasingly becoming 24/7 functions where customers expect to make purchases and transactions smoothly any time of day or night. Rather than disaster recovery, these organisations need a business continuity (BC) system that functions largely automatically so that customers are unaffected.
Continuity in the Cloud
Operating in a Cloud environment makes this feasible; it’s no coincidence that one of the biggest public Clouds is run by one of the world’s giant online retailers, Amazon.
Hybrid Clouds are increasingly common as organisations with on-premise data centres expand into the public Cloud rather than extending their building or relocating. It’s a quick and cost effective solution, but it leaves the business open to problems if disaster strikes the in-house part of the hybrid Cloud. Especially if some business critical parts of the system run entirely in-house.
There’s no point having a great Cloud-based BC system if it needs someone in-house to implement it when the data centre has gone offline.
Mike Curtis, CIO & COO, at Linux systems management specialist LinuxIT, defines three common mistakes in BC planning:
- Leaving out some of the systems and data that are business critical.
- Seeing some systems as business critical when they’re not.
- Over-engineering; the BC plan should be appropriate to the size of the business and the risks involved.
FCAPS (Fault, Configuration, Accounting, Performance & Security) is the ISO Telecommunications Management Network model now adopted as a framework for good systems management.
The ‘Fault’ protocols are of particular relevance to BC, and the most important aspect is the consideration of Reliability, Availability and Survivability (RAS) quality assurance. This aims to ensure that important IT services remain available in the event of component, system or even datacentre failure. Typically RAS provides enterprise-class data backup or replication, clustering of critical servers and components to ensure high availability. Some degree of service duplication, possibly with a second data centre or public Cloud vendor, enables the enterprise to keep trading even when the once-in-a-hundred-years disaster strikes.
When an eCommerce system is working well it can be difficult for management and IT to fully understand just what it would take to seriously disrupt the business, and to assess the risks.
It is possible to outsource disaster recovery, but it will require some due diligence to ensure that your organisation’s prized assets, including its data, are safe.