This past week if you are involved or use any kind of IT, you most likely heard about Amazon and Sony suffering a widespread outage. Amazon’s problem appeared to be a self-inflicted failure of maintenance and Sony’s appears to be an external attack on their network. Regardless the results were the same, their massive Cloud services went down and took many top tier applications with them.
Amazon is the leader on offering cheap, fast, powerful computing services through their Elastic Compute Cloud (EC2) service. If you have a credit card you can get a powerful server up and running, managed fully by Amazon to run whatever you need. Foursquare, HootSuite, and Reddit are some big names that use this service. All all those names went down and there was not much they could do about it other than wait and rely on Amazon to get their cloud services back up and running. An outage that had performance impacts over the span of days.
This is the dark side of the “Cloud” and turning over all your technology infrastructure into the hands of a service. You are losing control. Now the main applications affected are niche, luxury applications that had no widespread impact to anyone other than themselves. Seriously, do you think the world suffered if Quora was down for a few hours or you couldn’t check in with Foursquare. Those are play things, fun little toys. But it won’t be so pleasurable and easily set aside if Google Apps has a major outage, or the new Office 365. Those services are used to drive business, schools, financial institutions. As the industry is pushing for this centralization outside the walls and control of your company, sure you will save money, but is the savings worth risking days being offline?
Service Level Agreement contracts are words on paper. So what? You may not pay your next months’ bill if you suffer two days down, but no SLA will protect your against a failure. The Amazon Cloud outage appears to be a bad patch or network upgrade that propagated down and around, something on the surface that looked innocent and tested out OK, but fat fingering happens.
The Cloud is a nice thing to save startups and non-technical businesses to get to the IT level they need to be without hiring the people and finding a place to store their equipment. That’s fantastic and if I were to spin something up I would use Amazon’s Cloud service because a year’s cost if far less than the time and money to build servers here at my house. I am not against the Cloud movement. I am vocal about the marketing push that it’s a one all solution for the future.
I use Hootsuite as my primary Twitter application on the PC. When they were down, before I knew they were impacted by the Amazon outage, I cursed them. I had no clue, but I didn’t care. As a customer, Hootsuite is responsible for keeping their systems up if they want to keep me as a user. Since they run their business in the Cloud, they are held accountable for something that is 100% out of their control. That’s the dark shadow over the Cloud. Accountability and control.
You replace control for speed and cost savings. But like anything IT, Security, or any service technology, you don’t think about it until it fails you.
End of Line.
Binary Blogger has spent 20 years in the Information Security space currently providing security solutions and evangelism to clients. From early web application programming, system administration, senior management to enterprise consulting I provide practical security analysis and solutions to help companies and individuals figure out HOW to be secure every day.