Cloud in 2012: Don’t be Down and Outage
by Mark Herbert, Founder & Business Development Director, intY
Thursday, February 16, 2012
While 2011 saw cloud come of age with increased business adoption, for many, it was characterised by a series of high profile outages, BlackBerry, Microsoft and Amazon all fell – the collective grumblings of customers becoming headlines the world over.
Despite this public reckoning, industry analysts like Gartner and IDC continue to see a bright future for cloud. And so they should. The fact that some of the biggest players experienced outages demonstrates the reality of the cloud; it’s not a hallucination, it’s made up of real servers, stored in physical data centres. Like all IT, it will always be subject to some form of disruption, or disaster. What sets the cloud aside, is the capability of providers to mitigate this risk, and get you back up and running.
Of course, if you procure a service, best practice stipulates an expectation of service levels. It pays to be realistic though, no service can ever achieve 100% uptime. Even your own in-house data centre is unlikely to do so (and though I’m sure few would admit, in most cases, the impact is harder if your own services go down). And if anyone claims their service is infallible, they’re having you on.
Organisations need to seek the service and provider that places clear SLAs on disaster recovery. Downtime is, in most cases, rare but when it does happen the question you’ll be asking is ‘how quickly can operations be restored?’
Firstly, don’t underestimate customer service levels - feeling confident that you have access to your service provider in the worst case scenario is key. Guaranteed SLAs that speak to round the clock support - online and telephone - is an important safety blanket.
Due diligence at the procurement stage is important too. Review a service provider’s capabilities with the same rigour you would any of your own internal processes or services. This should include an assessment of your own DR capabilities, and what role you’ll need to play (not even cloud services mean you can absolve all responsibility!) in resuming services.
Consider, for instance, ‘premium’ versus ‘standard’, ‘gold’ versus ‘silver’ – every service provider has a classification. Buying the cheapest often means compromising on SLAs: customers who buy premium, on the other hand, will receive a priority service.
It’s unlikely, however, that all your services will require a priority DR response (and price), so identify which applications or data will need to resume service fast. This will enable you to better manage not only the restoration of services and resources, but also your budget.
Take mission-critical services such as internet, email and voice access. Any service provider worth their salt should understand the impact downtime can have here i.e. it could mean their customers cease trading. So, the premium DR package is what you are looking for here – comprehensive archiving and recovery, with back-ups stored offsite in secure data centres and maximum service availability.
Let’s be realistic, there will always be some events we can’t see coming - natural disasters, freak weather and the like. What you can do though, is make sure your business is best placed to survive it. Your agreement with your cloud provider is the lynchpin to making sure the right practices are in place to make the aftermath painless.
Like the analysts’ fearless and bold predictions for cloud in 2012 and beyond, I’ll make one too… There will be cloud outages. However, if it happens to you, it really doesn’t have to impact your business productivity.