While discussing things of the Internet, somebody once asked me, “Do I need a cloud?”
This question showed me once again how misunderstood the entire “cloud” concept is as regards the Internet, misunderstood even by so-called experts who really should know better. Further evidence appeared last week when part of the Amazon “cloud” crashed.
When it comes to the Internet term “cloud,” there isn’t just one “the cloud.” Instead, there are many clouds, run by many different companies and organizations. “Cloud” is simply a term for a group of computers that stores data and programs for people to access and use over the Internet. “Clouds” can make some things possible that may have been out of reach, before.
Rather than your company purchasing and maintaining large numbers of hard drives to store mountains of data, you simply rent “cloud” storage space and put your stuff there. You don’t even have to own expensive and powerful software, computers and/or servers to provide enterprise-level computing power to your employees; you just rent time from a cloud service, and use their expensive and powerful software, computers, and servers. All of your stuff “lives in the cloud,” instead of your own hard drive. What you give up in control, security and possession, you reap back as you become a more powerful and capable company. It’s all wonderfully hunky-dory until the unthinkable happens, and it all comes crashing down.
Many companies own and rent access to cloud services. Microsoft has a commercial cloud service called Azure, and a personal cloud service called OneDrive. Apple has a cloud service called iCloud. IBM has IBM Cloud. Others, like OpenStack, Carbonite, Google, VMWare, Citrix, Intuit, SalesForce and AT&T offer data storage, business management programs, productivity applications and computing power far beyond the reach of mere mortal computer users.
Unfortunately, all of these shiny and alluring cloud services suffer the same affliction: they are designed and built by imperfect people, using imperfect computers, in an imperfect world. Sometimes, when I think about it too much, I am genuinely distressed that we have based our entire society on such horribly flawed technologies as computers and the Internet. 30 years ago, when a computer crashed, it was bad, but it wasn’t a cataclysmic event affecting millions of people. 30 years ago, if a network crashed, sure, it was a bump in the road, but we had fairly stable systems in place we could fall back on, and life continued more or less as normal.
Not so, today. Now, when Internet systems and networks crash, most of us have nothing to fall back on. We are done. Toast. Game over. Back to the Stone Age. It’s as if an EF5 tornado ripped through town, but all it did was destroy the computers. Even though everything else is OK, what are you going to do? Think of all the things you do every day, as part of your normal life, that depend on a working computer, cell phone and Internet connection. That’s what happened last week to 124,000 websites and millions of Internet users, as part of the worlds biggest cloud service, Amazon Web Services, crashed.
Unknown to most Internet users, Amazon is more than just the world’s biggest online retailer; they are also the worlds biggest cloud service provider. Bigger than Microsoft, bigger than Google, Amazon Web Services sells essential Internet services to millions of websites, both great and small.
The 124,000 affected websites included Internet titans like Business Insider, Netflix, Reddit, the Associated Press and Adobe, who were knocked offline for as long as 11 hours. Countless “downstream” victims were rendered helpless, such as from Netflix, which accounts for 30% of all streaming video on the Internet. Thousands of “lesser” websites also suffered. College students were quoted by Computerworld as saying, “It’s knocked out my school’s technology back end. Students are freaking out because they can’t access assignments.”
The moral of this sad situation is a simple one, an old moral, one seemingly lost on the modernities of today’s allegedly sophisticated, tech-savvy populace: don’t put all your eggs in the same basket.
Yet, too many eggs were, in fact, in the same basket. When it comes to computers and the Internet, you’re supposed to have your eggs distributed across multiple, different baskets. That way, when one basket crashes (as it surely will), it doesn’t take your entire business down with it. You have backup eggs, which you press into service, and business carries on. This is fundamental stuff that I was taught years ago in network and security classes, before so many people placed so much blind faith in such flawed technology.
As pointed out by Computerworld’s Steven Vaughan-Nichols, small companies may have a financial excuse for not placing their eggs in more than one basket, as it can be somewhat expensive to keep your eggs in multiple Amazon cloud locations. Even so, there are affordable disaster recovery plans that can help any business avoid such crippling downtime. Companies like Netflix, however, whose downtime represented many millions of lost dollars, are without excuse. Rest assured, were it up to me, their business continuity folks would never work in my I.T. town, again.