Shadowrocket Node Timeout? Troubleshoot Latency Tests, Protocol Settings, and System Time

Seeing timeout or connection errors in Shadowrocket? Check the latency test, then verify the server address, port, password, and transport settings. Finally, check your device’s clock.

At a glance

A timeout means a test or connection stage did not finish within the allotted time. It does not necessarily mean the password is wrong or the server is down. This guide is for users who already have a provider and subscription. Keep test conditions consistent, then check protocol settings, routing mode, and system time. The steps use Home, Config, and Settings on iPhone and iPad as examples. Mac (Apple silicon and Intel), Windows, Android, and Linux users can follow the same troubleshooting sequence; control locations vary by device.

1. Tell a node test timeout from a website connection failure

The latency shown by Shadowrocket is the result of one test method on the network at that moment; it does not guarantee access to everything. ICMP, TCP, and HTTP tests take different paths. A server may block ICMP; a successful TCP handshake does not prove proxy authentication works; and HTTP results can depend on the destination, TLS, and the application’s response. Before comparing results, check the test method under Settings → Connectivity Test.

Do not compare millisecond results from different test methods. And a test that returns a value does not replace a real connection check. First note the network, test method, node, and stage where the error occurs. Change one thing at a time so you can identify what made a difference.

What you seeWhat it tells youNext step
ICMP times out, TCP returns a resultThe server may simply not respond to ICMP; this alone does not mean the proxy is unavailable.Keep the same node and test an actual proxy connection.
TCP times outA TCP connection to the destination port was not established in time.Check the address and port, then test again on another network.
TCP returns a result, but the website failsThe port is reachable, but authentication, TLS, transport, or routing may still be failing.Check the protocol settings and the policy actually used by the request.
Only the subscription update times outThe failed request may be the subscription update, not a saved node.Test an existing node and the subscription update separately.

Bottom line: keep the test method consistent

Use repeated failures under the same network, node, and test method as evidence for changing the configuration. Retest a single timeout first; do not delete the node right away.

2. Follow a consistent first-pass checklist

Start with steps you can easily undo. While troubleshooting, do not update the subscription, switch nodes, and edit Config at the same time. If the connection recovers, you otherwise will not know which change helped. These steps apply to a node already saved on your device. Get subscription links from your provider.

  1. Note the current setup

    In Home, note the selected node and current Global Routing mode. Then check the method under Settings → Connectivity Test. Record whether the test itself times out or a request fails after connecting.

  2. Check the local network

    Temporarily disconnect and confirm that the device can open pages that normally work without a proxy. On public Wi-Fi, complete any sign-in or captive portal step first. Do not change proxy settings until the basic network connection works.

  3. Run the test again

    Go to Home → Connectivity Test and run the same test three times under the same conditions. Treat three timeouts differently from one intermittent timeout; do not mistake a brief fluctuation for a persistent problem.

  4. Switch networks

    Keep the node and its settings unchanged, switch from Wi-Fi to cellular, and run the same test. If it fails on only one network, check that network’s port restrictions, UDP reachability, or authentication status first.

  5. Check automatic connection

    Open Settings → On Demand and note the current setting. If you need to control the connection manually while troubleshooting, temporarily turn off On Demand. Restore it when testing is done so automatic connection rules do not affect your observations.

“Three times” is just a convenient way to record results, not the app’s timeout threshold or a rating of connection quality. If the same node keeps failing on both networks, check its settings next. If other existing nodes under the same subscription work, ask your provider to verify that node’s listening status and account permissions.

3. Check the address, port, and protocol settings

In Home, open the target node’s edit screen and compare each field with the original details from your provider. The address is usually a hostname or IP; do not paste a full web address with a path into this field. The port must match the port the server is actually listening on. For example, if the details specify 443, do not change it to 8443 just because the connection fails. Port 443 alone does not mean the service uses TLS.

CheckMust matchCommon pitfall
Address and portThe hostname or IP and port provided by your service provider.Including spaces when copying, or using the subscription URL as the server address.
ShadowsocksEncryption method, password, and any plugin settings required by the server.Updating the password but not the encryption method.
VMess / VLESSUUID, protocol-specific authentication or Flow settings, and the actual transport method.Reusing fields from one protocol for a different protocol.
TrojanPassword and TLS settings, especially the name required by the server.Connecting by IP and leaving out the correct Server Name.
WebSocket / gRPCThe WebSocket Host and Path, or the gRPC Service Name.Using Path as the Server Name, or missing the correct capitalization or leading slash.
Hysteria2Authentication details, TLS name, and any obfuscation settings configured on the server.Using a TCP-only test to judge a QUIC connection that runs over UDP.
WireGuardSettings such as the key, Endpoint, port, and tunnel addresses.Treating a general TCP test as the result of a UDP tunnel handshake.

The TLS Server Name, WebSocket Host, and server address may be the same or different; do not change them to match just for the sake of consistency. If the server uses security settings that require an additional key or identifier, check each value against the original details. For subscription-generated nodes, save a record before updating so your manual changes are not overwritten by the next update.

4. If the connection is on, check routing and configuration

A VPN status in the system only means the system tunnel is in the corresponding state; it does not mean every remote request succeeds. If a node test returns a result but websites still fail, check Global Routing in Home. Proxy, Direct, Config, and Scene are different routing modes, not node protocols or network quality ratings.

With Config, rules are evaluated in order, and the matching rule determines the policy. Check whether an earlier DOMAIN-SUFFIX, IP-CIDR, or GEOIP rule sends the request to DIRECT or REJECT; FINAL handles requests that did not match an earlier rule. Check the request log rather than assuming traffic uses the proxy just because PROXY appears in the configuration file.

If only one domain fails, distinguish a DNS resolution failure from a proxy connection failure. If the log shows that the destination address was not resolved, check the DNS settings in the active configuration and whether the resolver is reachable. Do not treat a DNS resolution error as a node port timeout. Restore the original routing mode after troubleshooting so temporary test settings do not remain in effect.

5. Use the error stage to narrow it down, then check system time

Log wording can vary by operating system and connection stage. Use the English snippets below only if they appear in the actual log; not every failure has the same cause. Save the full error line and the time it occurred, but redact credentials before sharing it.

Error: timeout

Cause: A stage did not finish within the time limit. This message alone cannot tell you whether the issue is the network, port, or handshake.

Fix: Use the test method and surrounding log entries to identify the failed stage, then retest the same node on another network.

Error: connection refused

Cause: The connection was explicitly refused. This commonly means nothing is listening on the destination port, or an intermediary actively rejected the connection.

Fix: Check the address and port first, then ask your provider to confirm the listening status. Repeatedly changing the password usually will not fix a failure at this stage.

Error: TLS handshake timeout

Cause: The TLS handshake did not complete in time. The issue may involve a blocked route, the server response, or handshake settings.

Fix: Check that TLS is enabled as required by the server and that the Server Name is correct, then compare results on another network.

Error: certificate has expired or is not yet valid

Cause: The certificate’s validity period does not match the device’s current time, or the server certificate has actually expired.

Fix: Correct the device’s date and time, then reconnect. If the same error persists, ask your provider to check the certificate’s validity period.

On iPhone and iPad, open system Settings → General → Date & Time. Check that Set Automatically is on and that the displayed date and time are correct. A different time zone label does not necessarily cause authentication to fail; what matters is the device’s actual clock. Do not change only the displayed time zone while overlooking an incorrect date.

After correcting the clock, disconnect and reconnect so the handshake runs again. TLS certificate validation and time-sensitive authentication such as VMess can be affected by a significantly incorrect clock, but not every timeout is caused by the time. If the clock was already accurate, continue troubleshooting using the network and settings evidence above.

6. Common questions about subscription updates and node timeouts

Should I delete all my nodes if a subscription update times out?

No. Keep your current nodes and check whether they connect, then verify that the subscription link is still valid. Get the link from your provider. If refreshing from Home still fails, note the specific subscription request error. Do not clear the list without first saving your settings.

Why won’t a video load even though the latency is low?

The millisecond result describes only that test; it does not indicate sustained throughput or prove that a particular destination is reachable. Check Global Routing and the selected node, then see which rule the request actually matches. If other pages work but one service does not, note the destination and error stage instead of focusing only on latency.

Wi-Fi times out but cellular works. Should I change the protocol?

Keep the protocol unchanged at first. If the same settings work on another network, that points to a difference in the network path. For UDP-based connections such as Hysteria2 and WireGuard, check UDP reachability in particular. Test another supported protocol only if your provider has supplied the necessary settings.

What information should I send my provider?

Include the time and time zone, device operating system, protocol, network type, results from three tests, and redacted error lines. Do not include passwords, UUIDs, private keys, full subscription URLs, or tokens from those URLs in public screenshots or posts.

Bottom line: let the evidence guide the next step

If the local network works, the connection fails on both networks, and the settings and clock are correct, send your provider redacted logs to check the server status. Randomly changing more settings at this point usually just adds variables.

Download Shadowrocket