Amazon's S3 package offers storage and allows web services to be run from their cloud infrastructure. Many large sites are heavily reliant on this service to survive and has made their own presence online sustainable where it otherwise wouldn't be. Huge sites such as SnapChat, Netflix, Giphy, Slack, Reddit, Medium, Nest and Imgur all use AWS (Amazon Web Services) and were among the 120,000 domains that were affected by a recent outage.
Most hosting providers offer 99.9% uptime but Amazon has gone as far to say its services are built for "99.999999999 percent durability".
The problem with services like S3 is that they position themselves so far up the chain and have such a large number of users which need them to survive, that when something goes wrong at the top, everything below it falls which is what we saw on 28.02.17. Although network outages happen to everyone, even the likes of Microsoft and Google, they're still a shock when they do. Aside from the 120k websites which are built upon S3, there are countless thousands of 3rd party sites which rely on data from AWS hosted sites which are then in turn also affected.
Its no different to when the DNS provider Dyn was attacked recently which again caused thousands of sites to become inaccessible including some giants such as Twitter, Paypal and Spotify. Or when CloudFlare have had IP ranges blocked in error forcing hundreds of sites offline which are on their network.
People expect 100% uptime in today's world and these providers are fundamental in helping smaller companies access huge computing power and resources, often without paying a penny but when outages like the one at Amazon happen, it highlights how fragile these systems are and how a handful of companies are essentially propping up the internet!