[UPDATE: As of late Saturday afternoon, Instagram appeared to have much of its functionality restored for most users.]
Instagram has been insta-tripping since Friday night and here's why: A powerful storm in northern Virginia took out the servers it runs on.
Maybe you're wondering how a storm on the East Coast killed access to your favorite photo sharing app, which is based in San Francisco. That's because Instagram, like many major web companies, uses an even bigger web company Amazon to host its traffic and data.
An enterprise cloud-computing product called Amazon Web Services (AWS) powers businesses in 190 countries worldwide "hundreds of thousands" of startups and mature companies, all told, according Amazon's website. Pinterest, Netflix and Heroku are among the notable sites and services you may be familiar with in addition to, of course, "your Instagram."
Some call AWS the Coke of the web hosting industry, so that gives even more of an idea of just how prevalent it is.
AWS operates out of Oregon and northern California in addition to northern Virginia, but only the Virginia facilities were affected in Friday's storm. Still, that's enough to have people wondering how they'll see filtered Saturday brunch photos, as well as enough to make "My Instagram" trend worldwide on Twitter for several hours. (Pinterest and Netflix were back up by Saturday morning.)
The outage to Instagram and other major sites shows that despite massive hype and momentum in the Internet world cloud computing isn't necessarily a magic solution for businesses data and IT needs.
Amazon began offering the AWS service in 2006. It lets businesses host apps and websites, backup and store data, and generally run their enterprise IT. The idea as with cloud computing in general is businesses can stay more nimble by accessing servers from a remote location and renting space and capacity as needed. Scaling becomes easier and companies without resources for dedicated server hardware don't have to worry about that side of the business until something like this happens.
One product AWS offers is called Elastic Load Balancing as part of its Elastic Compute Cloud (EC2) service. This means a business can pay to set up EC2 service in multiple areas. Then, when service in a particular availability area isn't working say, because of a huge storm traffic is automatically rerouted to another, healthier availability area. Then load balancing is restored once all availability areas are healthy.
But if a company say, for example, Instagram doesn't have Elastic Load Balancing set up and has all of its IT needs hosted in just one AWS service area say, for example, Northern Virginia and a problem occurs, then it's a more serious issue. What's unclear is whether Instagram put all its data eggs in one AWS basket and in part brought Friday and Saturday's extended downtime upon itself, or whether AWS is failing Instagram entirely.
Mashable has contacted Instagram for comment, but so far not received a response. Amazon tells us it will release more information about the outages "in the coming days."
Alex Hazlett contributed reporting to this post.
No hay comentarios:
Publicar un comentario