ERR_CONNECTION_TIMED_OUT is Chrome's way of saying it asked for a connection and got silence — no answer, no refusal, nothing — until the clock ran out. The usual suspects are an unreachable host, a firewall silently dropping packets, or a mistyped proxy endpoint. Here's how to isolate the failing hop in minutes and fix the configuration mistakes that cause it.
ERR_CONNECTION_TIMED_OUT (Chrome network error -118) fires when the browser sends a TCP connection request and receives no response of any kind before the timer expires — typically around 30 seconds. That distinguishes it from ERR_CONNECTION_REFUSED, where the host actively rejects the connection: a refusal means something is listening and said no; a timeout means your packets disappeared into a void.
The silence has three usual explanations: the host is down or unreachable, the IP or port is wrong, or a firewall along the path is dropping packets silently instead of rejecting them. DNS is not the issue here — a failed lookup produces ERR_NAME_NOT_RESOLVED instead — so the name resolved to an address; the address just didn't answer.
If a proxy is configured, the first hop changes: the browser connects to the proxy before it ever contacts the target. A mistyped proxy hostname, a wrong port, or a corporate firewall blocking outbound connections to that port can all time out before the target site is involved at all — which is why the error can appear on every site at once.
Work from the most common cause outward: confirm the destination is actually reachable, then check each hop between you and it. Each step below isolates one layer of the path.
curl -v --connect-timeout 10 https://example.com/ shows exactly where the connection stalls.curl -x http://user:pass@proxyhost:port https://example.com/. A timeout here means the proxy endpoint, not the target, is unreachable.ipconfig /flushdns on Windows, sudo dscacheutil -flushcache on macOS) in case a stale record points at a dead address.With a proxy in play, ERR_CONNECTION_TIMED_OUT usually means the client-to-proxy leg failed, not the proxy-to-target leg — the browser gives up before the target is ever contacted. The usual configuration mistakes: a mistyped endpoint, a network that blocks the proxy's port outbound, or an authentication setup that no longer matches the machine you're connecting from.
Once the proxy connection itself is healthy, timeouts shift to the exit side: a residential exit on a slow or dying home connection can stall a request long enough to hit your client timeout. That's a rotation problem, not a configuration problem — retrying on a fresh exit IP resolves it. Setup guides for common tools live under /integration/ if you want a known-good baseline config.
Double-check the exact hostname and port from your provider dashboard, and verify your network allows outbound connections to it. ProxyOmega keeps this simple: one endpoint, residential.proxyomega.com:10000, and the same port speaks HTTP, HTTPS, and SOCKS5 — there is only one host-and-port pair to get right.
On rotating residential pools, an occasional exit will sit on a weak connection. ProxyOmega's Budget Unlimited (/plan-unlimited/) rotates each port on an interval automatically, and you can force a fresh exit at any time by changing the -session- ID in your username — so a retry leaves through a different IP instead of the same stalled one.
If you authenticate by IP whitelist instead of username and password, a changed egress IP — a new office connection, a redeployed cloud server — means the proxy rejects your connections. That usually surfaces as an immediate refusal or reset rather than a slow timeout, but it belongs on the checklist whenever every request through the proxy fails at once.
Update the whitelist in the dashboard, or switch to username/password auth (the password is your dashboard API key) so authentication follows you across machines.
Set explicit connect and read timeouts so a stalled connection surfaces as a catchable exception in seconds instead of a hung script, and retry on failure — with a rotating pool, the retry lands on a different exit IP.
import requests
PROXY = "http://USERNAME:[email protected]:10000"
proxies = {"http": PROXY, "https": PROXY}
for attempt in range(3):
try:
resp = requests.get(
"https://httpbin.org/ip",
proxies=proxies,
timeout=(10, 30), # 10s to connect, 30s to read
)
print(resp.status_code, resp.text)
break
except requests.exceptions.ConnectTimeout:
print(f"connect timeout on attempt {attempt + 1}, retrying")curl -x http://user:pass@host:port https://example.com/ — if that times out, fix the endpoint or the firewall, not the site.ProxyOmega serves HTTP, HTTPS, and SOCKS5 on one endpoint, with 1.5M+ rotating residential IPs so every retry lands on a fresh exit.
ProxyOmega 90M+ ethically-sourced IPs across 200+ countries and 30,000+ cities. Residential, mobile, ISP and IPv6 proxies for scraping and AI agents.