Last Tuesday, a variety of different websites and apps like Netflix, Reddit, Trello, Slack, and Pinterest were reported to be suffering from outage problems. Here at Howard Tech Advisors, some of our own services were down during the day. Luckily, we are in the business of IT security and had a backup plan in place for such issues. The problem? The Amazon Web Services platform, in particular the S3 cloud storage system, was having trouble. What happened? Why did it happen? What can you do about it now? In today’s blog, we’ll be exploring these questions.
In total, 54 of the largest online retailers had performance issues and many of the web’s most popular apps and services were down completely. Even Amazon’s own site relies on the S3 cloud storage system and the site they use to report updates and outages to the system was down temporarily. In total, companies lost about $150 million from the four-hour outage. Commentators around the internet observed the similarity between the Amazon Web Services outage and the cyberattack on Dyn servers in October, although the big difference between the two is that the Dyn server attacks were intentional.
Why Did it Happen?
According to Amazon, the problem in the S3 system originated due to an accidental typo by a team member who was attempting to debug their billing system. This typo removed more servers from the billing system than intended, and restarting the entire system was the only way to fix the problem.
On a macro level, this problem stems from the increasing consolidation of cloud services into a few companies. Amazon won’t say how large its cloud is, but it was estimated in 2012 that Amazon hosted around 1% of the entire internet in its cloud services. So an attack or accident on one of their services can wreak havoc on a wide variety of seemingly disparate websites and apps.
What Can Your Company Do Now?
If your company relies on Amazon Web Services and experienced a slowdown or outage last week, you might be wondering what to do next. Amazon’s cloud services are extremely popular and are increasingly controlling the infrastructure of the internet (and the economy as a whole). If you don’t have cyber insurance already, now’s a good time to look into it. While it may make sense for some companies to move workloads to the public cloud, you might also benefit from diversifying your approach and moving some of these workloads to private clouds or legacy systems to reduce your risk. If you experience similar issues, you can also check Amazon Web Services’ Service Health Dashboard.
HOWARD TECH ADVISORS: YOUR PARTNER IN TECH
At Howard Tech Advisors, we manage your IT infrastructure so that you don’t have to. Whether you need assistance creating a disaster recovery plan or you’d like to outsource your IT needs, we can help! Keep up with our weekly blog to stay up-to-date on the latest tech trends, security information you need to know to stay safe online, and tips and tricks to effectively navigating an increasingly mobile world.