Security, cost and reliability top concerns over putting data in the cloud, according to a recent survey commissioned by NeverFail. Consequently, just over one-third (37.6 percent) were unsure about the viability of using the cloud for disaster recovery and almost the same proportion (31.3 percent) rule out cloud for DR altogether.
NeverFail is one of many vendors encouraging enterprises to use its products in the cloud. It makes software that forces a server to fail over to a second, identical device should it fail. You would probably want to do this as part of your disaster recovery (DR) plan. The fail-over process applies to the software environment so the mirrors need not be physically identical, only logically so -- in other words, the backup can be a virtual machine. This might lead you to think that using the cloud for fail-over purposes is a natural next step but, according to the survey, there is, unsurprisingly, a lot of reluctance to do so.
The survey involved over 1,000 IT managers and C-level executives in the UK, working in both public and private organisation. Most of them were not NeverFail's customers, according to Bob Roudebush, NeverFail's VP of marketing.
I suspect Roudebush reckons he's mounting a pincer movement on potential customers when he went on to talk to me about how many companies now run mission-critical applications in virtual machines -- this is data gleaned from another section of the survey. According to him, 58 percent of the organisations surveyed are virtualising applications such as Microsoft Exchange, SQL Server, Sharepoint, Blackberry Enterprise Server, especially in smaller environments such as branch offices.
You can see what's coming: virtual machines need protection too and since there's more of them running critical applications, they need to be part of your DR plan -- which incidentally is needed not so much to protect against big but rare events such as fire and flood but against hardware and software failure such as failed cooling and power, and of course human error, which Roudebush reckoned accounts for about 15-20 percent of downtime; this puts it "about third" on the list of causes.
Other data from the survey includes concern over the cost of downtime. NeverFail reckons that 17.4 percent of businesses said that the hourly cost of downtime is over £5,000, while 47.5 percent said they didn't know the hourly cost of downtime. And 18.8 percent of IT staff have experienced downtime lasting more than a full business day and 85.2 percent of organisations had suffered an outage at some stage.
None of this is earth-shatteringly newsworthy, nor is it likely to be much that you couldn't have estimated for yourself, but I found it useful as a snapshot of how virtualised the UK's servers are, of cloudiness in general and because, on the face of it, it involves a statistically valid dataset.