OpenAI, Claude and Grok down in same hour yesterday. Where is your fallback?
Yesterday around 14:50 UTC people on HN notice ChatGPT, Claude and Grok all down at once: https://news.ycombinator.com/item?id=49551096
Claude status page: https://status.claude.com/incidents/461yvfrzpwtt. Elevated errors from 13:26 UTC, resolved 16:16 UTC. Affected list is basically every model they sell: Fable/Mythos 5.1, Fable/Mythos 5, Opus 5, 4.8, 4.6. And claude.ai, the API, Claude Code, Cowork. Everything.
Nobody has said why three companies fall over in same hour. Maybe coincidence. Maybe shared provider. Someone on HN pointed out Astra launched same day. I don't know. Is not the point.
The point: I ask five teams this morning "what happened to your product at 14:00 UTC yesterday". Four say "it was down". One says "it switched to local model, users saw slower answers". One out of five.
Three hours. Every provider. If your agent has no fallback, your product has no fallback. Is not hard. Is one more config line and one boring model on a box somewhere.
2 comments
Join the discussion
Log in to comment.
The garage rack was fine, for the record. Slow, dumb, and up. Reckon that's the pitch for local now: not cheaper, not smarter, just there.
Claude Code affected is the part that hurt me. I was mid-refactor, 12 files, and it just stop. Three hours, I go back to typing like a caveman. First time in months. Humbling, haha. Fallback config is now on my list.