Published: 20.09.2026
Every website goes down sometimes. What separates a small hiccup from a disaster is how fast you notice and how well you are prepared. Knowing the usual causes helps you prevent most of them.
The most common causes
- Expired domain or certificate. It is the most avoidable cause of all: a forgotten renewal.
- Traffic peaks. A campaign or a viral post exhausts CPU, memory or connections.
- Software errors. A faulty update or plugin breaks the application.
- Database problems. A full disk, a crashed table or too many connections.
- Server or network failures. Hardware, power or a network issue at the provider.
- Attacks. DDoS floods or hacks that change or remove content.
- Human mistakes. A wrong DNS record, a deleted file or a bad configuration change.
Monitoring: know before your customers do
Use an external monitoring service that requests your site every minute from several places. Check the home page and also important functions such as the login, search and checkout. Receive alerts by e-mail, messenger or phone, and set a short threshold so that you are not woken up by a single failed request.
What to monitor
- Availability and response time of key pages.
- Certificate and domain expiry dates with reminders weeks in advance.
- Server resources: CPU, memory, disk space and load.
- Error logs of the web server and the application.
- Backup jobs: a failed backup is a hidden problem.
Prevention
- Turn on auto-renewal for domains and certificates and keep payment details valid.
- Test updates on a copy before the live site.
- Use caching and a CDN so that peaks are absorbed.
- Choose resources with a reserve and scale before you hit the limit.
- Protect against attacks with a firewall, rate limiting and a proxy.
- Remove single points of failure for critical systems.
When it happens: a calm checklist
- Confirm that the site is really down from more than one place.
- Check the domain, DNS, certificate and the status page of your provider.
- Look at the server load, disk space and error logs.
- Undo the last change if the problem started after it.
- Contact support with the time, the symptoms and what you already checked.
- Communicate with customers on social media or a status page.
After the incident
Write down what happened, how long it lasted and how it was fixed. Decide what to change so that it does not repeat: a new alert, a bigger plan, a better backup. Small notes after each incident build a much stronger system over time.
Backups and recovery time
Know how long a restore takes and practise it. A backup that takes a day to restore may be too slow for a shop. Keep one recent copy in a different location.
Frequently asked questions
Is 100% uptime possible? No system can promise it. Aim for a high level and a fast recovery.
Is free monitoring enough? For a small site, yes. Choose a tool with several locations and instant alerts.
Does downtime hurt SEO? Short outages usually do not, but long or repeated ones can reduce crawling and rankings.
