5 comments

  • karmakaze 20 hours ago

    Funny, a telephone book is a good analogy--but some would need that explained.

    > DNS is like an internet location engine, converting user-friendly web addresses like amazon.com into IP addresses - a series of numbers that other websites and applications can understand.

  • hdgvhicv 19 hours ago

    All eggs in one basket. The same basket. Because of groupthink.

    Throw in penny pinching and cost cutting and you run on the edge. Nobody gets a bonus for spending money on backup systems. Nobody gets fired for buying ibm.

    • SOLAR_FIELDS 19 hours ago

      Part of this is also naïveté of people spinning up systems in AWS us-east-1. Any SRE or operator will tell you to never spin up systems in there unless you have a requirement to do so. It’s long been a tribal knowledge of the instability. This outage is more widespread than others but I see region level outages, at least partial ones, happen there every 1-2 years on average

      • hdgvhicv 9 hours ago

        Our file deliveries to from s3 failed because despite being eu-west-1/2 authentication relied on us-east-1

        From what understand that’s a feature you can’t opt out of.

        AWS is a single on if failure. The same point used globally by an industry that can’t imagine any other way.

        Diversity is defence. Monoculture is a strategic weakness.

  • ChrisArchitect 19 hours ago