TECH NEWS – The AWS outage brutally showcased the internet’s fragility.

According to Ars Technica, Monday’s disruption was astonishingly caused by a single software bug that rippled through AWS systems. The fault was traced to components tied to DynamoDB in AWS’s DNS management stack. DNS (Domain Name System) is often likened to the internet’s phone book: it translates human-readable domain names into the IP addresses machines use.

On today’s internet, many services (cloud platforms, streaming, and more) must map a single domain to multiple IP addresses to leverage geographically distributed servers efficiently. DNS Enactor—the component responsible for updating these records in DynamoDB—experienced unusual latency and had to retry updates across multiple DNS endpoints. While that Enactor was catching up, DynamoDB kept generating new plans, which a separate, on-time DNS Enactor began executing.

When the delayed Enactor finally caught up, it overwrote the newly created DNS configuration with a significantly older plan—also slipping past a safeguard intended to prevent exactly this kind of error (that safeguard was itself delayed). The timely Enactor then detected the stale plan and deleted it. The cascading behavior impacted AWS broadly, and engineers had to manually diagnose and remediate the issue.

It’s yet another reminder of how brittle the internet can be and how sensitive it is to subtle internal logic failures. A “simple” bug can knock down entire ecosystems—few things are more maddening than when everything should work on paper, yet doesn’t, and hardware isn’t to blame.

Source: PCGamer, Ars Technica