Overcoming Disaster

Overcoming Disaster

archives2013

By Neil Cresswell, CEO, Virtus Data Centres

Research just published by Zerto (who surveyed IT managers, VMware admins, System admins, plus business continuity and disaster recovery experts) found that 42 per cent had suffered an outage in the last six months. It also found that a combination of power loss/interruption and hardware failure accounted for two thirds of all outages.

The survey, (reproduced by permission of Zerto below,) quite correctly concludes that all businesses need a working disaster recovery (DR) process.

The first of the two main causes of downtime that the research identified as being responsible for 31.5 per cent of outages – was power loss or interruption. Although the research doesn’t drill down into the type of IT installation in these cases, it would be reasonable to conclude that most of these occurred in company’s own server rooms. The reason is fairly simple – most enterprises simply cannot cost justify the cost of deployment to the level of redundancy that is available and deployed at major Tier III colocation data centres.

At a major colocation provider like Virtus, the economies of scale mean that it is easy for us to provide dual power feeds from the national grid – plus multiple power and cooling distribution paths. Then redundant and concurrently maintainable UPS (Uninterruptible Power Systems), generators and other components -– are all available and able to withstand at least 72 hours of failure of the national grid supply or any other component.

In terms of the other major factor identified by the survey – hardware failure – there are considerable advantages to using ‘cloud’ providers rather than operating your own hardware. Many of Virtus’ customers are cloud providers and offer solutions with in-built redundancy at the hardware level. Whilst these don’t stop hardware failures – that would defy the laws of physics – they do stop outages due to hardware failures because their IT infrastructure is designed to be fault-tolerant.

This extra level of resilience is available whether the cloud provider is selling IaaS (Infrastructure as a Service) PaaS (Platform as a Service) or SaaS (Software as a Service).

So, as part of the disaster recovery plan recommended by Zerto as a conclusion from their research, two thirds of the problems can be very simply mitigated by switching from in-house server rooms to cloud-style computing provision in high efficiency, well staffed, tightly managed and high reliability Tier III data centres.

We can proudly say that since Virtus opened for business in March 2011 we have had a completely solid 100 per cent up time.

The Zerto research is reproduced here by kind permission.