A blank page, "unable to connect to proxy server," or an HTTP error raises one question: why is my proxy not working? It sits between your device and the website. The common causes are local, network-side, on the proxy server, or at the target site. A refused connection or instant timeout often means a wrong port, protocol, firewall rule, or unreachable endpoint. A repeated HTTP error usually narrows the search.
For example, if a marketer sees ERR_PROXY_CONNECTION_FAILED while a colleague uses the same details normally, the cause is likely local, not a dead proxy. Use this symptom list to pick the likely cause. Check one detail, retest, then move to the next fix before changing anything else.
Check the network path before changing settings.
For example, office Wi-Fi can block non-standard ports. The same details may work at once over a phone hotspot. That points to the local path, not a dead server.
If pages load only without the proxy, test the details in a free online proxy checker. If pages fail both ways, fix your local network first.
Most failed proxy connections start with settings that do not match the issued details. Check the IP address, port, HTTP, and SOCKS5 carefully. A wrong protocol choice can make the proxy time out silently. Copy each value from your proxy settings, especially the port, and remove trailing spaces before saving.
HTTP proxy and SOCKS5 proxy modes are not interchangeable. If you enter one type into the wrong field, requests fail or time out. Compare these proxy settings character by character with the provider’s details.
For example, a SOCKS5 port typed into a field labelled “HTTP proxy” can make every page hang. Such settings mismatches usually produce a generic connection error.
--proxy-server is ignored. Quit it fully or add --user-data-dir. For example, launching it while Chrome runs changes nothing. A separate profile fixes the proxy.407 Proxy Authentication Required means your request reached the proxy server, but it rejected or never received valid credentials. Although this error belongs to client-side error codes, the proxy side can still trigger it by expecting different details.
Typical causes include a typo in the username or an expired or replaced password. Hidden copy-paste characters, credentials tied to the wrong proxy, or an app that does not send authentication can also cause it.
In a browser, this error often looks like a sign-in prompt that keeps reappearing even after you submit the password.
If every test fails, the proxy server may be at fault.
An “unable to connect to proxy server” error message, ERR_PROXY_CONNECTION_FAILED, or a timeout often means the connection fails before the server responds. Test this proxy connection error in order: firewall or antivirus, browser extensions, then cached browser data.
Security software can intercept or block outbound proxy traffic on an unusual port. Check whether the device firewall or antivirus is stopping the connection.
For example, after an update, antivirus may silently drop a scraper's connection to the proxy port, and an allow rule restores it.
VPN, ad-blocker, privacy, and proxy-switcher extensions can override proxy settings or conflict silently. For example, a VPN extension may take over, so traffic never uses the intended proxy.
If it works only with extensions off, the last enabled one is the cause.
Old redirects, PAC data, cookies, or a cached error page can keep a failed state alive after the proxy details are correct. For example, if a website still shows “proxy refused connection,” clear the browser cache, then run the test once again.
DNS problems can look like proxy failures. A site may show “server not found” or fail only while the proxy is on. A DNS error means a hostname is not becoming an IP address on your network. A refused proxy connection is a different error. The host or port rejects the request.
For example, if “This site can’t be reached” appears only with the proxy on, switching resolvers and flushing the cache can restore the page.
Code | What it means | Most likely cause | First fix |
|---|---|---|---|
407 | Auth failed | Bad login | Retype |
403 | Denied | Blocked IP | Change IP |
429 | Throttled | Rate limits | Slow down |
500 | Internal fault | App fault | Retry later |
502 | Bad gateway | Upstream fault | Check proxy |
503 | Unavailable | Overload | Wait |
504 | Gateway timeout | Upstream slow | Test proxy |
HTTP and HTTPS status codes separate request faults from server faults. These proxy error messages narrow the cause. For a scraper, 429 means slow down. A 502 means check the proxy.
A target website may block a proxy IP address over reputation, prior abuse, datacenter origin, or suspicious behaviour. For example, a retail website returns 403 to a proxy IP but loads normally on a phone's mobile data.
Signs:
Test another network, slow requests, rotate the IP if allowed, and use normal headers. Understanding the Proxy Ecosystem (2020) notes many sites blacklist proxies.
429 Too Many Requests means the website’s rate limits were exceeded from this proxy IP. The tempting fix is to retry instantly because the task looks stalled. That pressure usually extends a short throttle into a longer block. Treat these rate limits as server-driven request limits and slow traffic:
500 internal server error, 502 bad gateway, 503 service unavailable and 504 gateway timeout are server error responses. They can come from the target website, your proxy, or an overloaded upstream.
Check provider status, then check announcements before changing settings.
If curl succeeds but the browser fails, the proxy itself is likely fine. A terminal request bypasses many local connection settings, so it gives a clean baseline. That isolates the error. The problem usually sits in the app’s proxy configuration.
curl -x http://user:pass@host:port https://example.com# Expected: page HTML or the exit IP
If that check works but the client fails, compare it with the working test:
For example, curl may return the exit IP, while a desktop app shows a connection error because it only supports passwordless servers. That message points to the client path, not the proxy.
Contact your provider when every check passes. If several sites still fail through the proxy server, the fault may sit upstream. Correct but rejected credentials, or a blocked IP address, also point to the provider. Before you contact support, repeat one check and note the exact proxy error. The provider may also confirm alternative ports.
Include these proxy details in the support ticket for a faster diagnosis.
For example, if one site returns 403 through the proxy, these details help the provider speed an IP change or replacement instead of guessing.
At OnlineProxy, mobile proxies run on real 4G/5G carrier IPs through modems and SIM routers. On a dedicated device, you can change the IP by link or timer, or reboot it when blocked. The provider offers 24/7 proxy support.
Check the host, port, and protocol first. Test the site without the proxy. Pause firewall or antivirus briefly, then disable browser extensions. Change only one item at a time, then retest one page.
This error means authentication failed. The proxy expects valid credentials but did not receive them. Re-enter username and password carefully. Remove copied spaces, then confirm they match the active port. Retest one request again.
Security tools may treat an unusual port or traffic as suspicious and break the connection. Allow the app, or pause protection briefly, then retest. Restore full coverage afterward. Open only one target page.
It is a three-digit response. Codes starting with 4 point to the request or client. Those starting with 5 indicate a server problem. Read the error code before changing anything. Then test again.
Usually the setting is off, a conflicting extension or VPN is active, or the details contain a typo. Retest after each correction. Use the correct protocol and port before saving it once again.
Turn it off. Clear the saved proxy settings. Restart the browser or device, then enter the details again and save. If it still fails, repeat the earlier steps. Then test one page again.