How Do You Survive in this Fragile Cloud?
Instagram, Amazon, Netflix hit by Virginia powerful storm. That is how fragile it all is. #InstaFails #Instagram http://t.co/Qen60UOG
— Joel (@joeltalks) June 30, 2012
For many administrators and end users, this past weekend was a perfect storm of web disasters. Web darlings like Yelp, Netflix, Instagram, Skype and Heroku were all down for at least a couple hours this weekend when a nasty storm in West Virginia caused a power outage that temporarily took down Amazon’s AWS Cloud Server System. Then just as companies started recovering, a Leap Second bug caused major havoc on websites such as LinkedIn, Reddit and Mozilla who weren’t prepared to account for the extra leap second transitioning from June 30 to July 1. This perfect storm meant some users couldn’t find a restaurant to eat at and others couldn’t upload photos to Instagram. But it also shed light on a more important truth: Many services we all enjoy and use on a daily basis that take advantage of the cloud’s “always on, stored anywhere” strength have still not set up the proper architecture for true redundancy.
Yeah, @FruzsE and I noticed. Bad Internet weather. RT @jorilallo: Today isn't a good day for Skype calls. You shouldn't even try
— ✍ Jon Mitchell ⚛ (@GoForArgus) July 2, 2012
While the rest of the technology world was down over the weekend, KAZOO users were running uninterrupted business as usual. Why? We’ve built KAZOO, keeping in mind that business communications simply cannot be compromised. KAZOO clients include ambulance dispatch centers and police stations – institutions that find it imperative to be up at all hours of the day, no matter what is happening elsewhere. Case study here.
KAZOO is based on the premise that systems will fail, and should fail, gracefully. We fully expect entire data centers to go offline, data to be lost or corrupted and networks to slow down or flap, and that’s only for starters. Storms happen, tornadoes come and go, and there are things we simply cannot predict no matter how technologically advanced we become. That’s why we’ve done what it turns out many other companies are just now starting to realize: the need for automatic failover redundancy.
Our architecture is designed so that even if a power outage like Friday’s were to take down our AWS servers, our backup clouds are ready on all different data centers and providers. (We <3 Rackspace & Linode) This is a huge, key differentiator of KAZOO - we can run on any Linux-compatible hosting provider (virtualized or dedicated) in the world. All we need is an SSH port to connect to.
In today’s world, the only way to survive in ‘cloud fragility’ is architecture built for live failover redundancy.
