· Troubleshooting · 8 min read
Clash Nodes All Timing Out? Troubleshooting Subscription, Proxy, and Connection Issues
Troubleshoot Clash node timeouts in order: refresh your subscription, check system time and ports, then verify system proxy settings, DNS, and firewall rules.
1. Refresh the subscription: verify the download and active configuration
In your client, open the Subscription or Profiles page, find the active profile, and refresh it manually. Check the update time and result, then make sure the refreshed profile is still selected. Seeing nodes in the list only means the client has retained node data; old cached entries may remain visible even when an update fails.
- HTTP 401 or 403: The subscription server rejected the request. Check your account status, whether the subscription has expired, and whether the link was reset. If the link has changed, import the new one instead of repeatedly refreshing the old URL.
- HTTP 200, but the configuration fails to parse: Check whether the client received a login page or service notice instead of a configuration, or whether the format is incompatible with your Clash core. Keep your last working configuration and obtain a compatible format from your subscription provider.
- Update succeeded, but some groups are empty: Check whether the referenced
proxy-providersloaded successfully on its own. Updating the main configuration and updating proxy providers are separate steps; check the logs for each.
2. System time: rule out TLS certificate validity errors
If the logs show certificate has expired or is not yet valid, check the system date, time, and time zone first. The error may occur while fetching an HTTPS subscription or connecting to a node that uses TLS. Check for clock drift after restoring a virtual machine snapshot or powering on a device that has been off for a long time.
- Windows 11: Open Settings → Time & language → Date & time, turn on Set time automatically, then select Sync now.
- macOS Ventura and later: Open System Settings → General → Date & Time and check the automatic time and time zone options.
- After syncing, retry the step that failed: refresh the subscription again if it failed, or retest the same node if its handshake failed. If the time is correct but the certificate error persists, check the server certificate and your device's trust chain. Don't mask the problem by disabling certificate verification.
3. Port conflicts: check whether mixed-port is actually listening
mixed-port is the local entry point shared by the HTTP and SOCKS proxies. The examples below use 7890; this is not a universal default across clients. Use the runtime port shown in your client. external-controller is the management interface, not the proxy port to enter in apps. A desktop client's port setting may also override the value in the subscription file.
Use the listening status to distinguish a port conflict from a core that hasn't started
In Windows PowerShell, query the listening port and its process ID:
Get-NetTCPConnection -State Listen -LocalPort 7890 |
Select-Object LocalAddress, LocalPort, OwningProcess
In macOS Terminal, see which process is listening:
lsof -nP -iTCP:7890 -sTCP:LISTEN
If nothing is listening, check whether the core has started, the configuration loaded successfully, and whether the active port differs. If the logs show address already in use, identify the process using the port. If it's another proxy client, quit that client normally. If another application needs the port, change the current client's port to one confirmed to be free, such as 7891, then update the system proxy setting. A listening port confirms that the local entry point is up; it doesn't mean a node is connected.
4. System proxy: compare a direct proxy request with your browser
Test the local proxy first, then check the system setting. Leave the client running, temporarily turn off TUN, and send an explicit request to a local proxy endpoint that supports HTTP. The example uses 7890; in Windows PowerShell, use curl.exe.
curl -v --proxy http://127.0.0.1:7890 --connect-timeout 5 --max-time 15 https://example.com/
- Connection to the local port is refused immediately: Recheck the listener. Verify the port, address, and protocol; in particular, don't use a SOCKS-only endpoint as an HTTP proxy.
- The local port connects, but the request fails: Check the client logs for the same time. Determine whether the target rule selected
DIRECT, the policy group didn't select the expected node, the node handshake failed, or the remote request timed out. Explicitly using the local proxy only proves that traffic reached the core; it doesn't guarantee the core will select a node. - The command returns a page, but the browser still fails: Turn the client's system proxy setting back on and check the address and port recorded by the operating system. On Windows 11, check Settings → Network & internet → Proxy. On macOS, check System Settings → Network → the current network service → Details → Proxies. Also check whether the browser's own settings, extensions, or organization policies override the system settings.
The system proxy only affects apps that honor proxy settings. If your browser works but one app doesn't, first check whether that app supports HTTP or SOCKS proxies. To test TUN, enable it separately and retest the same app. Avoid changing the proxy setting, routes, and DNS all at once, or it will be difficult to tell what caused the difference.
5. DNS: distinguish node hostnames from destination hostnames
If the logs show no such host or a DNS timeout, identify the hostname in the error. If it matches the node's server value, the failure occurs before the node connection is established. If it's the website you're trying to reach, consider the matched rule and outbound route to determine whether DNS is handled locally or remotely. These two failures need different troubleshooting steps—not the same generic advice to change DNS.
- Check system DNS resolution. Use
nslookupto query the hostname from the error and note the DNS server and result. A successful system lookup only shows that the system's default DNS path works; it doesn't prove the core can reach its configured upstream. - Check core DNS resolution. In mihomo,
dns.proxy-server-nameservercan specify a resolver for node hostnames. If it isn't configured, node hostnames may use the general DNS settings. Check the effective configuration and make sure the upstream is reachable before the proxy is established. For older cores, first confirm that the field is supported. - Check for TUN-specific behavior. If connections work with TUN off but fail when it's on, check TUN's
dns-hijacksetting and the route to the DNS upstream. Make sure DNS requests aren't being sent through a proxy route that depends on the same DNS result. Withfake-ip, a query handled by the core may return a synthetic address from its configured range; that's expected and doesn't indicate a DNS failure.
6. Firewall and network: compare connection logs for the same node
If the node hostname resolves but the logs keep showing connection timeouts, note the node address, port, transport protocol, and failure time. Switch to another network you're authorized to use, such as a mobile hotspot, and test again with the same device, configuration, node, and test URL. If it only fails on the original network, first check its sign-in requirements, outbound restrictions, or endpoint security policies. If it fails on both, ask your service provider to confirm the node is online.
For TCP-based nodes, use Test-NetConnection in Windows PowerShell to check TCP connectivity to the actual node address and port. A successful TCP connection only confirms that a connection was established; it doesn't verify TLS, authentication, or the proxy protocol. For UDP-only transports, a TCP test says nothing about the port's UDP status. If security software logs a block, follow your organization's policy to request the necessary outbound permission for the confirmed client or core process. Don't disable your entire security suite just to test.
Retest in this order
First confirm that the subscription updates successfully, then verify that a node can handle a real request. Finally, restore your usual rules mode, system proxy, or TUN settings and retest each change. If real websites load while the dashboard continues to show timeouts, check the test URL, response status, and timeout threshold separately. When reporting an issue, include the client and core versions, failure time, and redacted logs. Remove the full subscription URL, tokens, and node credentials.