"The site is down" covers four very different failures with four different owners. These four commands, in order, place it - and each one is quick.

1. Does the name resolve?

dig yourdomain.com +short

No answer means DNS: the domain, the nameservers or the record. Nothing at the server can fix it, and restarting things there is wasted time.

2. Does the address answer?

ping -c 3 1.2.3.4\ncurl -sI --max-time 10 http://1.2.3.4/

Ping refused but curl answering is normal - many providers block ICMP. Neither answering means the machine is off or the firewall is closed.

3. Is the web server running?

systemctl status nginx\nss -tulpn | grep -E ":80|:443"

Nothing listening on 443 with a working DNS record is the whole answer.

4. What does it actually return?

curl -sI https://yourdomain.com/ | head -1

A status code here hands the problem to the right article: 502 is PHP, 504 is a timeout, 403 is permissions or a rule, 500 is the application.

Is it only you?

Before anything else, check from a phone on mobile data. A local DNS cache, a VPN or an ISP block affects one person and looks exactly like an outage - and half of reported outages are this.

If everything above is fine

  • A CDN in front, failing while the origin is healthy.
  • A certificate that expired - the site loads over http and refuses over https.
  • A firewall rule that blocked one country, or your own address after too many login attempts.
Write down what you find at each step. Handing "dig answers, port 443 is closed" to support is a fix in minutes; handing over "the site is down" is a conversation.