The problem rippled across the internet affecting everything from Cloud-based storage to mobile applications, and brought down some of the world’s most popular online services – including Netflix and Amazon’s own Prime Video System.
Issues were reportedly so widespread that Amazon could not even access their own AWS dashboard to notify users of the problems.
Businesses took to social media to advise users that services were down due to the outage. Meanwhile users took to twitter to lambast Amazon for the failure, with a number of humorous angles, including many pointing to the irony of the site isitdownrightnow.com being brought offline.
What was the cause?
The underlying cause of the outage was attributed to inaccessible systems in the US East 1 region which is located in Northern Virginia. Screenshots of the AWS Event Log Dashboard have emerged showing a number of logs identifying operational issues in this region.
Efficacy of the Cloud
Whilst many calm-headed commentators have remarked that Cloud outages have become less frequent and diverted some blame onto users for failing to co-locate. The sheer scale of this issue has posed some serious questions on the efficacy of public Cloud for critical systems.
A look at AWS’ client list will show the sheer breadth of organisations that can be affected anytime a problem occurs. This particular outage is reported to have affected the US Securities and Exchange Commission as well as Apple, Expedia and Snapchat. Another name on the AWS client list is the Central Intelligence Agency.
Too Big to Fail
Whilst the long list of high profile organisations is a testament to AWS’ success in cornering the market, the result of so much global infrastructure being reliant on a single organisation heaps huge responsibility and economic importance upon their shoulders.
The economic impact of this latest outage is yet to be calculated, but the scale of the disruption begs the question of whether a single organisation should have such a critical influence that it can bring global services to their knees.
Has AWS become too big to fail?