Check if the site is actually down
First, make sure the problem is not on your end. Open the site from your phone using mobile data, not Wi-Fi. Ask a colleague or friend to open it from a different location.
If the site loads for others, the problem is with your internet connection, router, or local settings. Restart your router, try a different browser, turn off your VPN.
If no one can access it, use a service like Down for Everyone or Just Me or Ping Admin. Enter your site address and it will show whether the site is accessible from different locations worldwide. Takes ten seconds and immediately tells you the scope of the problem.
Check hosting and domain payment
The most common reason a site goes down is an expired hosting account or domain registration. Hosting providers usually send a warning a week ahead, but the email is easy to miss.
Log into your hosting control panel and check the balance. If the service is suspended, top up the account - the site will be back up within an hour. Also check the domain status through a whois service: if the domain has expired, you need to renew it through the registrar.
This is nothing to be embarrassed about and happens even to large companies. But if you delegated payment to a contractor and they forgot, it is time to reconsider who is responsible for infrastructure.
Look at what the browser shows
Error messages on screen are clues. 502 Bad Gateway or 503 Service Unavailable means the server is overloaded or not responding. This happens during a sudden traffic spike or an attack. Restarting the server through the hosting panel often helps.
500 Internal Server Error means an error in the site code. Perhaps someone updated a plugin or changed a configuration. If you have server access, check the error logs - they will say which file or script broke.
DNS_PROBE_FINISHED_NXDOMAIN means the browser cannot find the site by name. Either the domain is not paid for, or DNS records were changed and have not propagated yet (this takes up to 24 hours). Check that the nameservers are correctly specified in the domain settings.
A blank white screen with no error text is usually a PHP or database problem. Turn on debug mode if you know how, or call the developer immediately.
Check your email and notifications
Hosting providers and monitoring services send emails when something goes wrong: disk space ran out, SSL certificate expired, server went down. Open your email and look for messages from the past 24 hours from hosting, domain registrar, or monitoring systems.
If there is an email, it usually says exactly what happened and what to do. For example, inode limit exceeded means there are too many files on the server - you need to clear cache, old backups, or logs.
Set up notifications on your phone if you have not already. That way you learn about problems before customers start complaining.
Roll back recent changes
If the site was working and then suddenly went down, recall what changed in the past few hours. Updated plugins? Edited code? Changed server settings?
Most hosting providers make automatic backups daily or weekly. Go to your hosting panel, find the Backups section, and restore a version of the site that definitely worked. Takes five minutes and often solves the problem without diagnosing the cause.
If files were changed directly on the server, check the modification date via FTP client. Delete or revert the changed file to its old version.
After restoring, make sure to figure out what exactly broke the site, or the situation will repeat. If it was a plugin update, postpone it and inform the developer.
Check server load and disk space
A site can crash due to lack of resources. Open the hosting panel and look at the graphs: CPU load, memory, disk space. If any of these are near 100%, that is your cause.
Disk space runs out due to bloated logs, backups, or uploaded files. Delete old archives, clear the log folder. If you do not know what is safe to delete, ask the developer or hosting support.
High CPU load can be due to a traffic spike, bots, or malicious scripts. Check your visitor statistics: if traffic increased several times, it is likely an attack or search engine crawlers. In such cases, connecting a CDN or temporarily restricting access via .htaccess helps.
Check the SSL certificate
If the browser shows a warning about an insecure connection, the problem is the SSL certificate. It may have expired, failed to renew automatically, or been installed incorrectly.
Open the site using http:// instead of https:// - if it loads that way, the certificate is definitely the issue. Go to the hosting panel, find the SSL section, and check the expiration date. Most hosting providers issue free Let's Encrypt certificates with auto-renewal, but sometimes this process breaks.
Reissue the certificate through the panel or contact hosting support - they usually resolve this in ten minutes.
When to call the developer and what to tell them
If the steps above did not help, it is time to bring in whoever built the site. But the more information you gather, the faster they will find the cause.
Tell the developer: when exactly the site went down, what the browser shows (error text or screenshot), what you already checked, whether there were any changes to the site in recent days. Provide hosting and domain access - without it they cannot do anything.
If the developer who built the site is no longer available, reach out to a new one. We at EFIMOV DEV regularly restore projects built by others and work with code we did not write. Diagnosis usually takes anywhere from half an hour to a few hours, depending on complexity.
How to avoid downtime in the future
Set up site monitoring. There are free services like UptimeRobot that check availability every five minutes and send a notification to Telegram or email if the site does not respond. You learn about problems before customers do.
Create a calendar for hosting and domain payments. Set a reminder a week before expiration. Or set up auto-payment if the hosting supports it.
Arrange maintenance with the developer or studio. It is inexpensive - usually a few hundred dollars a month - and solves 90% of problems before the site goes down. A package includes updates, backup checks, minor fixes, and standby for emergencies.
Regularly test your backups. It is not enough for them to just run - you need to try restoring them at least once every six months. Otherwise, at the critical moment you will find out the archive is corrupted.
Most site accessibility problems are solved in five minutes if you know where to look. Check the hosting balance, open the site from another device, look at the error message and your email - often the cause is obvious. If that does not help, gather all the information and pass it to the developer. To reduce these situations, set up monitoring, a payment calendar, and regular maintenance.
Frequently asked
The site does not open only for me - what should I do?
Check your internet connection, restart the router, try a different browser or device. Open the site using mobile data. If the site works everywhere except on your computer, the problem is with local settings, browser cache, or DNS.
How long does it take to restore a crashed site?
Depends on the cause. If the problem is unpaid hosting, up to an hour after payment. If it is code or configuration, from half an hour to several hours. Rolling back from a backup takes 5-15 minutes.
Can I restore the site myself without a developer?
Yes, if the cause is simple: unpaid hosting, full disk, need to restart the server, or roll back from backup. All of this is done through the hosting panel. But if there is an error in the code or database, you need a developer.
What is site monitoring and why do I need it?
It is a service that constantly checks whether your site is accessible and sends a notification if it does not respond. You learn about the problem immediately, not when customers start complaining. There are free options like UptimeRobot.