Platform dependency keeps biting small builders
A thread says Docker pull fails in Spain because Cloudflare is caught up in a La Liga football block. Commenters say the block takes out everything proxied through Cloudflare, including R2, and Docker Hub is collateral. A separate AWS thread shows a billing glitch with people reporting estimated bills in the hundreds of millions and even billions on accounts that normally cost under $10 a month. Older threads resurface the same fear: Stripe shutting down a business, Google Cloud suspending production projects at 1am on a Saturday, Slack closing its IRC and XMPP gateways, and Namecheap terminating Russian customers.
The pattern: one vendor decision, legal block or billing bug can cut you off, and you usually hear about it from users or on HN before the vendor's status page.
Some commenters push multi-provider architecture. Others point out that this is expensive and that most small teams cannot afford it.
So what?
Map your single points of failure: payments, DNS/CDN, registry, cloud account. Set billing alerts and keep a documented escape route for the two or three that would kill you, even if it is only a cold backup you've tested once.
Read these
Tell HN: Docker pull fails in Spain due to football Cloudflare block
AWS: Inaccurate Estimated Billing Data – $1.7 billion
Stripe has decided to nuke my entire business
Tell HN: Google Cloud suspended our production projects at 1am on Saturday
Tell HN: Slack decides to close down IRC and XMPP gateways
Namecheap: Russia Service Termination