In my country we have saying: elephant falls, grass suffers. But in hosting, one elephant falls and thousand customer lose data.
Question for community: how many of you still use single provider for critical infrastructure? Not asking about backup. Asking about architecture psychology.
I see same pattern after every incident. Two weeks of "we learned", then back to cheapest option with best specs. Human nature or market failure?
(the funny thing about "never again" (which everyone says (literally everyone (I have logs))) is that it's actually "never again until the price is right again")) (and the price is always right again eventually because (and this is the part that kills me) redundancy costs money that CFOs don't want to spend until after the fire))
(also "critical infrastructure" means different things to different people (to my last employer it meant "the thing we forgot was single-homed until it wasn't"))
This thread has been noted for its valuable firsthand perspective.
As a general reminder: our community guidelines discourage speculation about specific unverified incidents involving identifiable facilities. The broader architectural discussion remains welcome.
For those building new infrastructure: we maintain a provider diversity checklist in our documentation section. No single recommendation—evaluate your own risk tolerance.
I keep three copies of everything that matters. Two providers, one offline. Cost is 40% more than single-provider setup. Sleep quality is 400% better. 🍜
The psychology is not about learning. It is about forgetting. Three years is long enough to forget the sound of a datacenter evacuation alarm. I do not want to remember it again.
Post a reply
You need an account to reply.
Log in or
register to join the conversation.