The site is 'down' — and every minute of downtime costs money, sales and search rankings. The good news: most outages aren't random, they have typical, predictable causes. Let's go through the eight most common ones and what to do about them in advance.
Half of all downtime isn't 'bad hosting' but an expired domain, a forgotten SSL or exhausted plan resources. All of that is under your control.
8 reasons a site goes down
- Resource overload: traffic or scripts ate the plan's CPU/RAM
- Expired domain: you forgot to renew and the site vanished from the web
- SSL certificate expired: browsers block access
- A DDoS attack with no protection on the host's side
- A bug in code or a CMS/plugin update
- Data-center-side problems (outage, network, power)
- Database limit or connection count exhausted
- Human factor: a bad deploy with no backup
How to cut downtime
- Set up external uptime monitoring (alerts within minutes)
- Enable auto-renewal for the domain and SSL
- Configure regular automated backups
- Keep resource headroom — don't run at 100% of the plan
A site almost never goes down 'suddenly'. There's almost always an earlier signal — it just wasn't being monitored.
Tophosting editorial
Bottom line
Site stability isn't luck, it's hygiene: monitoring, auto-renewal, backups and resource headroom. But if outages keep happening because of the host (frequent data-center failures, weak support), that's a reason to switch providers. In the Tophosting catalog you can compare hosts by real reviews and our own speed tests.
FAQ
How do I learn the site is down before customers do?
Set up external uptime monitoring (UptimeRobot and similar) — it checks the site every minute and instantly alerts you by email or messenger.
Why does the site go down under a traffic surge?
Most often the plan's resources run out (CPU, RAM, DB connections). The fix is resource headroom, caching, or moving to a scalable plan.
Who's to blame for downtime — me or the host?
It depends on the cause. An expired domain/SSL and code bugs are on you; data-center and network failures are on the host. Logs and monitoring show exactly which.
