All our eggs in one cloud: When AWS sneezed & the Internet caught a cold

The recent service outage that Amazon Web Services (AWS) experienced in the US brought several dependent services to a standstill throughout the world. The outage, brought on by a domain name system (DNS) resolution failure that affected a crucial database service, leading to a cascading effect on AWS services. Dependent services that were impacted by this outage include Disney+, Duolingo, Fortnite, HBO Max, Robinhood, Roblox, Snapchat, Slack, Venmo, and Zoom. This also included a host of Amazon-owned operations including its main retail site and the Ring doorbell company.

As per Downdetector, a site that observes internet outages, over 2,000 companies were affected globally, with 8.1 M reports of problems from users including 1.9 M reports in the US, 1 M in the UK and 418,000 in Australia.

Though many of the sites were back within a few hours, some faced problems the entire day. By Monday evening, Amazon said that all of its cloud services had “returned to normal operations”.

But the incident that rattled investors about long term viability has experts questioning whether it’s wise to depend on just a few big providers for so much. Amazon, Microsoft and Google play a key role in the cloud market, AWS being the world’s largest cloud computing platform.

“We urgently need diversification in cloud computing. The infrastructure underpinning democratic discourse, independent journalism and secure communications cannot be dependent on a handful of companies.” — Dr. Corinne Cath-Speth, the head of digital at human rights organisation Article 19

The Guardian quoted Dr. Corinne Cath-Speth, the head of digital at human rights organisation Article 19, “We urgently need diversification in cloud computing. The infrastructure underpinning democratic discourse, independent journalism and secure communications cannot be dependent on a handful of companies.”

Sam Barker, VP Market Research, says, the geographical reach of cloud services offered increases the complexity of network architectures, “The larger the cloud service becomes, there is more within the service to fail, and a larger chance for cascading failures. The event serves as a reminder that even leading cloud providers are not immune to large-scale disruptions.”

Having said that, outages are inevitable because no system can be foolproof.

“Outages such as this are a clear reminder that no service provider, regardless of reputation, scale or reach, can guarantee 100% uptime or provide complete protection against service downtime,” — Sam Barker, VP Market Research

“Outages of any service are inevitable at some point. Outages such as this are a clear reminder that no service provider, regardless of reputation, scale or reach, can guarantee 100% uptime or provide complete protection against service downtime,” says Barker.

He advises that organizations cannot completely rely on services from AWS, and other leading providers like Google and Microsoft, especially those operating across multiple regions. Multi-cloud strategies to increase redundancy and avoid vendor lock-in are the way to go. Also, organizations should have their own monitoring software. Though this does mean an increase in spend in cloud platforms.

This means platforms that can offer reduced spend and investment into integrating with additional systems will increase their value proposition to enterprises.