Hub works on primary broadband but goes offline during failover. A partial failure matters: one capability still functions, while a closely related step does not. This guide focuses on isolating that difference in wi-fi and routers, not on indiscriminate resets. Use the known-good behavior as your control, change one variable at a time, and retest the original scenario after each intervention.
Practical six-point diagnosis
- Reproduce: Hub works on primary broadband but goes offline during failover.
- Compare: Test hub DNS resolution after WAN failover while phone remains online.
- Check: Hub retains stale gateway or DNS cache, without adjusting unrelated settings.
- Check separately: Failover VLAN policy excludes IoT.
- Correct: Confirm failover subnet and allowed outbound DNS/HTTPS then reboot only hub when safe.
- Verify: Hub reconnects automatically after primary/backup WAN changes.
Baseline test: Test hub DNS resolution after WAN failover while phone remains online. Record the current device model, OS or firmware version, and whether this occurs with one file, account, network or source. A reproducible contrast is more informative than a vague message such as “not working.” If the control fails too, stop treating this as a feature-specific problem and investigate the shared dependency first.
Possible explanations
| Hypothesis | Evidence worth checking | Next decision |
|---|---|---|
| Hub retains stale gateway or DNS cache | Compare network association, address and gateway information on the affected device with a known-working device. | If the control still works, focus on the affected path rather than resetting the whole device. |
| Failover VLAN policy excludes IoT | Compare network association, address and gateway information on the affected device with a known-working device. | If this condition is not present, leave the associated setting unchanged. |
| Cloud reconnect timeout | Repeat the same action under a controlled alternate state and record the exact difference in outcome. | Only escalate to this branch if the simpler comparison did not explain the failure. |
Choose the appropriate route
Route A — Correct the observable mismatch
Confirm failover subnet and allowed outbound DNS/HTTPS then reboot only hub when safe. Start only after capturing the baseline. When the recommended control is not present on your model, look for its supported equivalent in the product documentation rather than following screenshots from another version.
Return immediately to the failing task and check whether the outcome changed. If it did, preserve the working setting and avoid piling on additional changes. If it did not, undo the change where appropriate before moving to deeper checks.
Route B — Test the underlying dependency
Inspect WAN failover events and DHCP lease duration without opening inbound access. This advanced comparison is worthwhile only if the simple check did not resolve the problem. The first two possibilities to separate are hub retains stale gateway or DNS cache and failover VLAN policy excludes IoT. Do not interpret their presence as proof of a diagnosis; a compatible device and a partially failed service may coexist.
Route C — Handle a limitation or persistent fault
A third explanation is cloud reconnect timeout. Check whether the feature requires particular hardware, account privileges, region settings or source formats. When a feature is unsupported, there may be no legitimate settings-only fix; use the vendor-documented alternative rather than an unsafe workaround.
Protect your setup: Do not expose hub's local control interface to internet. Keep the current SSID, encryption, DNS and router configuration documented. Use a trusted network and never weaken firewall or WPA settings just to make a test pass.
Step-by-Step Troubleshooting
Follow the Beginner steps first. Advanced steps are for isolating hardware, application or network-specific causes after a safe baseline test.
Confirm the boundary
Hub works on primary broadband but goes offline during failover. Do not assume this is a complete outage. Test hub DNS resolution after WAN failover while phone remains online. Keep the account and connections unchanged between attempts.
Capture a useful baseline
Write down the affected model, software version, active input and when this started. Test hub DNS resolution after WAN failover while phone remains online. A working comparison tells you which part of the chain is still healthy.
Use the least disruptive correction
Confirm failover subnet and allowed outbound DNS/HTTPS then reboot only hub when safe. Apply one supported change at a time; stop if your model uses a different setting or does not advertise the feature.
Confirm the expected outcome
Hub reconnects automatically after primary/backup WAN changes. Try the original failing action again, not an unrelated easier test; retain the old setting until you know the change helped.
Test the first dependency
Investigate whether hub retains stale gateway or DNS cache. Specifically, compare network association, address and gateway information on the affected device with a known-working device. Compare the result with the working baseline from the beginner route.
Separate another possible cause
Assess whether failover VLAN policy excludes IoT. For this branch, compare network association, address and gateway information on the affected device with a known-working device. Do not change that configuration unless the evidence supports it.
Inspect the third branch
Inspect WAN failover events and DHCP lease duration without opening inbound access. Also consider cloud reconnect timeout. Record the exact point of failure and check product-specific support limits.
Validate and plan a rollback
Hub reconnects automatically after primary/backup WAN changes. Do not expose hub's local control interface to internet. If the last change made matters worse, restore it and request supported assistance with a redacted troubleshooting summary.
Verify the result
Success looks like this: Hub reconnects automatically after primary/backup WAN changes. Run this check on the same device and account that originally failed. A different device passing is useful as a comparison, but it does not prove the original issue is fixed.
What to check if it still fails
Before contacting support, collect the exact model and version, the outcome of the baseline test, and whether the controlled correction changed anything. Keep passwords, private media, account identifiers, serial numbers and network keys out of screenshots or shared logs. Do not expose hub's local control interface to internet.
Solution Checklist
After the fix
Repeat the original action once under normal conditions and once after a routine reconnect or reopening, if relevant. Hub reconnects automatically after primary/backup WAN changes. Keep a brief note of the app, input, account or firmware state that changed so you can recognize a recurrence.