Website Down? What to Do First
When a site is down, confirm first that it fails for everyone and not just your network. Then read the exact error: a DNS failure, a 500, a 502, a 503 or a billing lock each has a different fix. Change one thing at a time, and keep a backup ready.
For website owners handling an outage who need to distinguish local access faults, service errors and recoverable application changes.
knowledge
- Read time
- 4 min
- Sections
- 7
- Updated
- 14 Nov 2026
- Written by
- WebJi® team
Last editorial update: 14 November 2026.
See whether it fails for everyone or just you
Test the unavailable website from another connection and device before deciding whether the fault is local or wider. Compare mobile data with Wi-Fi and a private window. Two failed checks do not prove that every visitor is affected; use independent observations as well.
If it fails on mobile data too, the site or its DNS is down for everyone, and the rest of this page applies. The public status page for this network is network status.
Read the error you actually see
Read the exact DNS, TLS or HTTP error before choosing a fix; similar-looking outages can have different causes. A 502 means an invalid upstream response, a 504 means an upstream timeout, and a 503 indicates temporary unavailability. The code alone does not identify the failing service.
| What you see | What it usually means | What to do next |
|---|---|---|
| DNS error, name does not resolve | Domain expired, or its A record or nameserver is wrong | Check the domain expiry date and the DNS record in the host panel |
| 403 | A file or folder permission, or a firewall rule | Undo the last permission change; ask the host if a rule was added |
| 404 on every page | The document root or a rewrite rule is wrong | Restore the .htaccess or the site root path from the last working copy |
| 500 | The application threw an error, often a plugin or a broken file | Read the error log, then disable the last plugin or theme change |
| 502 or 504 | The web server waited on PHP or the database and gave up | Check PHP workers, database connections and slow queries |
| 503 | Overload, maintenance or a resource limit | Read CPU and entry processes in the panel |
| Database connection error | The database service is down, or its user or password changed | Check the database service and the credentials in the config file |
| SSL warning or expired certificate | The certificate lapsed or the domain on it changed | Renew or reissue the certificate after the DNS is correct |
Undo the last change
- Note the plugin, theme, deploy or file you changed last, with its time.
- Turn that one change off and purge the cache.
- Reload on mobile data, not on the browser tab you already have open.
- If the site returns, leave the change off until you know why it broke.
Undo a recent website change when its timing and logs make it a plausible cause, preserving data needed for recovery. A rollback may not safely reverse database migrations or new customer transactions. Test the same failing action after reverting the selected change.
Check the host panel, DNS and SSL certificate next
Check account notices, domain expiry, authoritative DNS and certificate validity when no recent application change explains the outage. Verify an unpaid invoice through the normal authenticated client area; the symptom alone does not establish a billing problem.
If the panel reports suspension, resolve the stated reason with the host. A working server does not rule out a domain or certificate problem; a resource-limit message needs diagnosis before buying a larger plan.
Restore a backup only when the files are the broken part
Restore when the files or the database are damaged and a rollback does not bring the site back. Copy the broken state first so the cause can still be found, then restore from a point before the break.
Check uploads and the database separately, because a backup can be partial: a site that loads with missing images has restored files but not the media folder. Test the restore on a staging copy when the host allows it, and only then replace the live site.
Review backup and recovery options before replacing production data. Use WordPress error logs to connect the failure to a specific change rather than restoring blindly.
Tell users and search engines while the site is down
For planned temporary downtime, use an appropriate 503 response and, where possible, an estimated Retry-After value. The HTTP status reference distinguishes temporary unavailability from a missing page. Tell users what is happening without exposing internal errors or credentials.
Pause paid campaigns while the site is down, otherwise you pay for clicks that land on an error. When the site is back, post the same one-line update in the same place, so nobody keeps retrying an old page.
What to check after the site is back
Confirm the fix held for an hour, write down the start time and the cause, and test that the backup you rely on actually restores. Then check whether the same fault can repeat before the next campaign; how to prevent a crash when traffic spikes covers the capacity side.
If the host must look, email support@webji.services. That mailbox is 24/7 in Hindi and English. The office number is for business hours and is not a support line.
Frequently asked questions
How do I know the website is really down?
Open it on mobile data with Wi-Fi off, then in a private window. If it loads there, the fault is your network, DNS cache or browser; if it fails there too, gather independent checks before assuming a global outage.
What does a 500 error mean?
The application on the server threw an error. A plugin, a theme or a broken file is the usual cause, so undo the last change first and reload. If the error started on its own, read the server error log for the failing file.
What if the website fails only for me?
Hard-refresh the page, clear the browser cache, and switch off a VPN or proxy. A site that opens on mobile data is up for everyone else, so the fault is on your network or device.
What do 502 and 503 mean?
A 502 means the web server got an invalid answer from PHP or another upstream service, and a 504 means that upstream took too long. A 503 means the server is overloaded or in maintenance, and a planned 503 with Retry-After tells search engines the downtime is temporary.
When do I restore a backup?
Restore when the files or the database are the broken part and undoing the last plugin does not bring the site back. Copy the broken state first, restore from a point before the break, and check uploads and the database separately.
Last editorial update: 14 November 2026.
Ready to get started with WebJi®?
Join 12,000+ businesses across India and overseas. INR pricing, 18% GST shown before payment and 24/7 email support in Hindi and English — average first reply under 60 minutes.