If your VPN connects but doesn’t work, check your public IP first, then DNS, and finally retest with the app that’s actually having trouble. “Connected” describes the client’s status; where a particular request goes also depends on the proxy mode, routing rules, and the app’s own network settings.
Compare your public IP before and after connecting
Your public IP is the address a website sees when it receives a request—not your device’s address on the local network. Keep the device, network, and browser unchanged during the test. Don’t switch Wi-Fi, routes, and browser settings at the same time, or it will be hard to tell what caused a different result.
- If your device allows you to briefly disconnect the VPN, pause sensitive activity first, then disconnect. Use the same HTTPS public IP lookup page to note the address. If it lists IPv4 and IPv6 separately, record each one. If your work device is managed by policy, don’t disable protection just to capture a baseline.
- Connect to the route you selected. Once the client status is stable, reload the lookup page in the same browser. Note the current route, proxy mode, and results. Don’t treat a result page that was already open before disconnecting as a new measurement.
- Compare the same address type. If the IPv4 address changes, the IPv4 egress used for that lookup has changed; this says nothing about IPv6 or other apps. If the original IPv6 address still appears after connecting, and your settings are meant to route both types of traffic, check how the client handles IPv6.
An IP address can change while its listed location stays the same; different exit addresses may be assigned to the same region, and IP geolocation databases may update at different times. Conversely, a changed location label alone doesn’t replace comparing public IP addresses. If the lookup page doesn’t detect a type of address, that doesn’t prove all traffic of that type is routed through the VPN.
Check DNS: distinguish DNS lookup results from your traffic’s exit point
DNS translates domain names into addresses. A DNS test page typically asks your browser to query a temporary domain, then displays the recursive resolvers that handled the query. This shows information about the resolver, not necessarily the DNS address configured directly on your device or the public IP used to access a website. Different locations alone don’t prove a DNS leak.
In the same browser, note the resolver name or provider before and after connecting, without changing the browser’s secure DNS settings. When secure DNS is enabled, the browser may send encrypted queries to a resolver it has selected; other apps may still use system DNS. Even if the same public resolver appears both times, that doesn’t tell you whether queries went directly or through the VPN route.
Check split tunneling and system proxy settings by app
A successful browser lookup only confirms that lookup worked. A system proxy usually affects only programs that follow the system proxy settings, and apps may have their own proxy settings. Virtual network adapter modes can route more traffic, but may still be affected by app exclusions, routing rules, and address type. Mode names vary by client, so don’t assume that “global” means all traffic is covered.
| What you see | Check first | How to retest |
|---|---|---|
| Lookup still shows the original public IP | The lookup domain may match a direct-connection rule, or the browser may not be using the current proxy. | Check the rule for that domain. If the client shows connection details, reload the page and check whether the request used a direct connection or the selected route. |
| Browser egress changes, but a desktop app still fails | The app may not follow the system proxy, or a routing rule may exclude it. | Check the app’s network settings, fully quit and restart it, then repeat the action that originally failed. |
| Only the target website says your region is unavailable | The target domain may follow a different rule. The site may also use your account region or its own IP location data. | Check which route the target request uses, then review the site’s regional rules and the exact error message. |
| No new websites open after connecting | DNS resolution may have failed, the route may not be connected properly, or proxy settings may conflict. | First note whether the error says the domain can’t be found, the proxy can’t connect, or the connection timed out. Then troubleshoot accordingly. |
When retesting, repeat the original action first—for example, refresh the sync status in the desktop app—instead of going back to the browser to check your IP again. Existing connections may have been established before the route changed; restarting the app can help test a new connection. If the client shows rule-match details, check both the target domain and the test domain. If they match different rules, the lookup result can’t confirm what happened to the target app.
How to tell: Verify an app that should use the VPN route with a new request from that app. If it matches the expected route but still errors, check the target service’s error message instead of assuming every failure means the VPN isn’t working.
Retest checklist and troubleshooting notes
- ✅ Change one thing at a time, such as switching routes or adjusting split tunneling. Keep the before-and-after results; don’t change DNS and proxy mode at the same time.
- ✅ Record the affected app, network type, route, mode, and exact error message. Note whether all requests fail or only a particular domain or action.
- ✅ Retest with the original settings on another network you’re authorized to use. If it fails only on the original network, first check whether you need to complete a captive portal sign-in or whether that network has an explicit access restriction.
- ❌ Don’t share subscription links, account credentials, or diagnostic logs containing access tokens publicly. Don’t disable certificate verification to deal with connection errors.
If the problem started only after switching routes, repeat the same action on the original route to see whether the difference is reproducible. If every route fails but regular websites open when disconnected, check the proxy settings for your platform in the setup guide, then continue troubleshooting by symptom in Troubleshooting. To get help, send redacted comparison results through the contact page. Instead of saying only “it connects but doesn’t work,” note which app failed and which route handled the test request to help narrow down the cause.