View all newsletters
Receive our newsletter - data, insights and analysis delivered to you
  1. Leadership
  2. Digital Transformation
October 13, 2015updated 17 Oct 2016 11:54am

Amazon’s cloud storm shows need for multiple providers

By John Oates

One of the benefits of cloud-based architecture is meant to be reliability and excellent uptime. But last week saw a major outage which brought down some of the internet’s biggest names including Airbnb, Netflix, Reddit and Tinder. All them were relying on the world’s biggest cloud provider.

Amazon’s Web Services (AWS) was down for about six hours last Sunday. The problem was metadata servers failing to answer queries quickly enough, so queries were resubmitted and the servers were quickly overwhelmed almost like an internal denial of service attack.

Cue red faces at Amazon, and at its customers with no plan B.

Of course this is not only Amazon’s problem – most of the other major cloud vendors have suffered similar problems in recent years.

We should stop thinking of ‘cloud’ versus ‘in-house’ as ‘reliable’ versus ‘unreliable’. Or indeed as ‘cheap’ versus ‘expensive’. It must be part of any decent IT portfolio but not all of it.

In fact uptime is not the most important reason to have more than one cloud provider.  Most businesses can survive the tiny outages a decent provider might still suffer.

But more than one provider, and the ability to truly switch seamlessly between them, will not only save you if there is a serious outage but also make negotiating contracts a much more pleasant process.

Content from our partners
Scan and deliver
GenAI cybersecurity: "A super-human analyst, with a brain the size of a planet."
Cloud, AI, and cyber security – highlights from DTX Manchester

You wouldn’t buy a server from someone who claimed to offer 100 per cent uptime. You should be just as suspicious of cloud vendors making similar promises.

And let’s not forget that server provision – whether it’s in your office, a distant data centre or a blue-chip Icelandic cloud provider – is just part of the mix.

You’re also relying on power supplies, telephone networks, transport availability, air conditioning systems and even staff’s ability to get to work.

As an aged technology hack I’ve spoken to data centre managers getting the fire brigade to spray water over their buildings in an effort to keep temperatures down when cooling systems failed. And I’ve spoken to others busy filling sand bags to stop flood water getting in.

Things can and will go wrong. As long as you’re not over-reliant on just one provider the chances are your business, and your job there, will survive.

Topics in this article :
Websites in our network
Select and enter your corporate email address Tech Monitor's research, insight and analysis examines the frontiers of digital transformation to help tech leaders navigate the future. Our Changelog newsletter delivers our best work to your inbox every week.
  • CIO
  • CTO
  • CISO
  • CSO
  • CFO
  • CDO
  • CEO
  • Architect Founder
  • MD
  • Director
  • Manager
  • Other
Visit our privacy policy for more information about our services, how Progressive Media Investments may use, process and share your personal data, including information on your rights in respect of your personal data and how you can unsubscribe from future marketing communications. Our services are intended for corporate subscribers and you warrant that the email address submitted is your corporate email address.