Developer infrastructure
- Is Auth0 down?Up
- Is Bitbucket down?Up
- Is CircleCI down?Up
- Is Clerk down?Up
- Is Cloudflare down?Up
- Is crates.io down?Up
- Is DigitalOcean down?Up
- Is Docker Hub down?Up
- Is GitHub down?Up
- Is Heroku down?Up
- Is Hugging Face down?Up
- Is Netlify down?Up
- Is npm down?Up
- Is PyPI down?Up
- Is Railway down?Up
- Is Render down?Up
- Is Resend down?Up
- Is RubyGems down?Up
- Is SendGrid down?Up
- Is Stripe down?Up
- Is Supabase down?Up
- Is Twilio down?Up
- Is Vercel down?Up
AI services
Developer tooling
General
How these pages work
Every 5 minutes, from Europe
Each service is requested every five minutes from our own infrastructure in Europe. We probe the endpoint the service actually serves (the registry, the API, the edge) and say on every page exactly which URL and status code we expect.
Failures are confirmed before we report them
When a check fails we immediately re-check from US East, then again from Europe 30 seconds later. Only when all three agree do we count a failure. One isolated failure is a blip; two or more in a row are an incident. If the confirming region is unreachable we say so rather than guessing.
Blocked is not down
Some sites answer automated requests with a bot challenge or a rate limit. We record those as blocked, never as outages: being unable to reach something is not the same as it being unavailable.
Uptime you can reproduce
Uptime is successful checks divided by counted checks over our scheduled samples. Downtime is the number of confirmed failed checks times five minutes. Gaps in our own coverage are drawn as gaps, not interpolated.
Checking your own site instead? Use the website down checker. Our probe identifies itself as DrumbeatsUptimeBot; vendors can opt out via robots.txt or email.
Monitor your own services the same way
Uptime checks from multiple regions, cron and background-job heartbeats, and alerts to Slack, email or webhooks. Free for up to 50 monitors.
Start monitoring free