Tangled Mess

Is this a love letter to Amazon’s cloud? Readers cry foul

TLDR: The piece argues today’s internet is so intertwined that old disaster recovery playbooks don’t work, pushing teams deeper into big cloud. Commenters clap back, calling it an Amazon apology tour and slamming multi-location setups as costly and hard—raising real stakes for businesses betting their uptime on the cloud.

The article paints a grim fairy tale: we used to run our own servers, now everything lives on someone else’s—and it’s all tangled together. The author’s big message: in a world of endless subscriptions and third‑party services, old-school “just flip to the backup site” fantasy is over. But the comments? Pure fire. One reader calls it a “cloud apologist” piece, accusing it of saying AWS is painful but inevitable—so stop dreaming about building your own safety net. Another blasts it as “terrible advice,” claiming Amazon makes running in multiple locations so hard and pricey that only giants like Netflix can swing it. For non‑tech folks: “multi‑region” means running your app in two distant places so one can rescue the other if things break. The crowd argues it’s either a luxury sport or a rigged game. Jokes fly about a “spaghetti cloud” where every app tugs on ten others, plus the classic “this is fine” dog sitting in a burning data center. The vibe is split: defenders say today’s internet is a web of dependencies we can’t fully control; detractors say that’s exactly why we should fight for simpler systems—or at least not crown the cloud king without a trial.

Key Points

  • Legacy on-premise colocation environments had frequent reliability issues, prompting investment in secondary sites and replication.
  • Vendor contracts increasingly required disaster recovery procedures and testing, driving compliance practices.
  • Cloud providers and open-source options enabled faster development with higher availability, fueling cloud migration.
  • Early cloud flakiness led organizations to attempt multi-region redundancy, with Netflix cited as a success example.
  • SaaS proliferation increased interdependencies, leading the article to claim traditional secondary-site redundancy is insufficient today.

Hottest takes

“cloud apologist post” — andrewstuart
“This is terrible advice” — outofmyshed
“Cross-region data transfers are too profitable for them” — outofmyshed
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.
Tangled Mess - Weaving News | Weaving News