Why Websites Go Down: The 7 Most Common Causes
From DNS misconfigurations to certificate expiry, DDoS attacks, and cloud-provider outages β the real reasons popular sites go dark, explained clearly.
Every major outage in the last decade β Facebook's six-hour blackout, Cloudflare's global cascade, AWS's US-East-1 saga β traces back to a handful of root causes. Here are the seven that matter most.
1. DNS misconfiguration
DNS translates domain names to IP addresses. When someone pushes a bad DNS change (a wrong record, a missing zone, an expired domain), users worldwide can't resolve the name β the servers themselves are still running, but nobody can find them. This is what took Facebook offline in October 2021.
2. Expired SSL certificates
Browsers refuse to connect to sites with expired TLS certificates. If nobody's watching cert expiry, one Saturday morning the whole site starts throwing scary security warnings and traffic drops to zero.
3. Cloud-provider outages
Most modern sites run on AWS, Google Cloud, Azure, or Cloudflare. When one of those providers has a region-wide incident, thousands of unrelated sites go down together. Look at our outages page during any major AWS event and you'll see the pattern.
4. Deployment failures
A bad deploy pushed to production can crash the app, corrupt a database migration, or introduce an infinite redirect loop. Good teams have automatic rollback; the rest scramble.
5. Traffic spikes
Getting posted to the front page of Hacker News, going viral on TikTok, or landing on a Super Bowl ad can bring 100Γ normal traffic in minutes. Sites without autoscaling or a proper CDN buckle under it.
6. DDoS attacks
Distributed denial-of-service attacks flood a server with fake traffic until it can't respond to real users. Mitigation (Cloudflare, AWS Shield, etc.) is standard now β but attacks scale too, and successful ones still happen.
7. Database problems
A locked table, an exhausted connection pool, a full disk, or a runaway query can take the whole site down while every other component works fine. Database issues are the sneakiest because monitoring often shows "web servers healthy" β but every request 500s.
How to tell which one it is
Third-party status pages, social media, and monitoring services like ours can usually narrow it down within an hour. Search for the service on Down For All Or Just Me β our probes and community reports often surface the pattern (DNS errors vs. slow responses vs. hard failures) before the company posts an update.
Related articles
Ready to check if a specific service is down right now? Search on the home page or see the live outages.