Tony Howlett

Mar 28, 2017

Is your AWS truly Resilient?

If you host critical company infrastructure solely on a public cloud like AWS or Azure, you might be wondering how resilient that infrastructure is. After a major AWS outage, their largest region (Virginia) was affected for hours a few weeks ago. Large organizations like SnapChat and NetFlix were affected, as were many small- to medium-sized companies. Your company may have decided that it makes financial sense to use the public cloud as a primary rather than a secondary (or tertiary) backup site. But when AWS experiences a major issue like this, your company can be the one left with a black eye in the public’s mind, regardless of who was at fault.

Considering the Benefits

Amazon’s cloud does offer impressive redundancy capabilities, like the ability to collocate systems and data in different zones around the world. They also distribute data and processing power across hundreds or even thousands of disks. They provide a level of fault tolerance that few companies could afford to emulate, with astronomical uptime percentages. But that doesn’t tell the full tale of what you can expect as a pure AWS cloud infrastructure company.

Stuff still manages to happen. While most outages are not as severe as the full region, multi-hour outage of a few weeks ago, minor services hiccups occur more often than Amazon would like to admit. There is no such thing as an unsinkable ship, and “failure of imagination” type outages are always a risk when all your eggs are in one basket, even one as well designed as Amazon’s.

Considering “Insurance”

Got core functions running solely on public cloud? Then it’s a good idea to make sure you fully utilize all of the AWS functionality to maximize uptime. But it might also be worth having a little AWS insurance policy: keep at least some of your most crucial infrastructure on an alternate hosting provider that has nothing to do with AWS. Cost of these services is low, especially if you are only hosting backup DNS, domain or webservers there. You may not be able to replicate 100% of your infrastructure for full functionality during an AWS outage, but at least you can continue to provide basic web and network services to your employees and be able to communicate with customers and the world to let them know what is happening. Make sure your AWS is as resilient as you need it to be.

Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedIn

Tags: , , , , , ,