How to check if a website is down (for you or for everyone)

You open a client's site and get a spinning loader, a timeout, or a blank page. Your first instinct is to escalate: message the host, call the client, start looking for a cause. But the right first move is to find out whether anyone else can reach it, because if they can, the outage is yours to solve locally.

How to check if a website is down is simpler than it sounds: visit downforeveryoneorjustme.com or isitdownrightnow.com, paste the URL, and read the result. If the tool says the site is up but you cannot reach it, the fault is on your end. If it says down, the site is unreachable from outside your network and the outage is real.

How to check if a website is down for everyone or just you

External checkers work by requesting the URL from their own servers, which sit outside your network and often in multiple regions. If the server returns a response, they report the site as up. If the request fails or times out, they report it as down.

That single fact determines what you do next. "Down for everyone" means the problem is at the server, the host, or the DNS provider. "Up for everyone but you" means the problem is between your device and the site, and the list of culprits is short.

These are the most common local causes, in rough order of likelihood.

Your browser cache is holding a stale error. Browsers sometimes cache error states alongside normal pages. Press Ctrl+Shift+R on Windows or Cmd+Shift+R on a Mac to force a full reload that bypasses the cache. Alternatively, open the site in a private or incognito window, which does not use cached data at all.

Your DNS cache has a stale record. DNS translates domain names to IP addresses, and your computer keeps a local copy of recent lookups. If the site recently moved to a new server and changed its IP address, your computer might still be sending requests to the old one. To flush the DNS cache:

• Windows: open a command prompt and run ipconfig /flushdns
• macOS: run sudo dscacheutil -flushcache; sudo killall -HUP mDNSResponder in Terminal
• Linux: run sudo systemd-resolve --flush-caches

After flushing, try the site again.

Your VPN is routing traffic through a broken path. VPNs intercept DNS queries and route your connection through their own servers. If the VPN is having issues, or if the site is geo-blocked in the country your VPN exits from, you will see a failure that no one else sees. Disconnect the VPN and try again. If the site loads, the VPN was the cause.

Your ISP has a routing problem. Less common, but it happens. If you have a mobile device on a different carrier, try loading the site over mobile data. If it loads there, your ISP's route to that server is the problem, not the site itself.

Your IP is blocked. Some hosts and CDNs block IP ranges after too many failed requests or as part of geographic restrictions. If you have been running tests against a server, your IP may be on a block list. This is rare but worth knowing.

What to do when the site really is down

If an external checker confirms the site is down for everyone, you have a real outage. A few checks help you understand what kind.

Try the site by IP address. If you know the server's IP (you can look it up with nslookup yourdomain.com or dig yourdomain.com), try accessing it directly in the browser as http://[ip address]. If it loads, the server is up but DNS is broken. If it times out, the server itself is unreachable.

Check the host's status page. Most hosting providers and CDNs maintain a public status page. If they are experiencing an incident, they usually post it there before your support ticket gets a response. A quick search for "[your host] status" finds it.

Check whether it is the whole domain or one page. Try the homepage and a subpage. If one path returns an error and another does not, you may have a broken deploy or a misconfigured route rather than a full outage.

Read the HTTP status code. A 503 means the server is up but refusing to handle requests, often because it is overloaded or in maintenance mode. A 502 means the web server is up but the application behind it is not responding. A timeout with no response at all usually means the server or the host's network is completely unreachable.

Who to tell, and when

Once you have confirmed a real outage, the question is who needs to know.

If you manage the hosting, you already have everything you need to act. If the client manages their own hosting, your practical options depend on whether you have access to the control panel.

What changes everything is knowing before anyone else does. When you have monitoring in place, you have a timestamp for when the outage started, you know its duration, and you can reach out to the client before they reach you. That is the difference between looking like you caught a problem and looking like you ignored one.

The worst version of this situation is not the outage itself. It is finding out from an angry client message, hours after the fact, with no data about when it started or how long it lasted. At that point you are explaining yourself rather than solving a problem.

For clients who want visibility into this data, a public status page turns your monitoring into something they can check themselves, which deflects a surprising number of "is it down?" messages. The full setup, from picking what to watch to getting alerts before anyone complains, is covered in how to monitor your clients' websites.

FAQ

How do I check if a website is down without an online tool?

Switch from your Wi-Fi connection to mobile data and try loading the site. If it loads on mobile data but not on Wi-Fi, the problem is local to your network or router. If it fails on both, the site is more likely down for everyone. You can also ask someone in a different city to try loading it.

Why does a website checker say a site is up when I cannot load it?

The checker is testing from its own servers, which take a different route to the site than you do. If the checker can reach it, the server is responding normally. The problem is somewhere between your device and the site: your DNS cache, your VPN, your ISP, or a firewall rule. Work through those in order.

What is the quickest first step when a client's site looks broken?

Open the site in an incognito or private window. This skips the browser cache entirely. If the site loads in incognito, your main browser had a stale cached error and a regular cache clear will fix it. If it still fails, paste the URL into an external checker to determine whether the outage is local or real.

How long should I wait before escalating a real outage?

If the site is confirmed down from outside your network, do not wait. Contact the host immediately and loop in the client if the outage has lasted more than a few minutes. Hosts generally respond faster when you open a ticket early, and clients appreciate being told rather than finding out themselves.

Can a site be partly down?

Yes. A server can be up and serving some requests while others fail. Common causes include a crashed application process, a database that stopped accepting connections, or a CDN edge node with a problem in one region. If the homepage loads but a checkout page or login returns an error, that is a partial failure. It still needs fixing, but the cause is different from a full outage.

Trrack.it monitors uptime, SSL and response time for all your client sites and alerts you the moment one breaks. Your first site is free, so you can get started in a couple of minutes.