Why Does Home Assistant Fail Only for One User or Device?
A Home Assistant failure limited to one user or device usually comes from account scope, stale client state, URL or certificate handling, DNS, VPN, or that device's network path.
Do not restart or reconfigure the working server first. Build a small matrix: test the affected account on a known-good client, a known-good account on the affected client, and the same client on Wi-Fi and cellular or another network. The first comparison that moves the failure identifies whether identity, client state, or transport deserves the next action.
Prove the Failure Scope With a Cross-Test
Use the same Home Assistant URL and dashboard for four comparisons: affected account on affected device, affected account on known-good device, known-good account on affected device, and known-good account on known-good device. Keep network and URL constant for the first round.
A community Android case recommends trying another browser to determine whether failure follows the browser or the device. That cross-browser discriminator is more informative than clearing every setting at once.
If the failure follows the account, continue to identity and permissions. If it follows the client, continue to cache, app, URL, and certificate tests. If it appears only on one network, skip client resets and inspect DNS, routing, or VPN policy.
Check Account Scope and Authentication State
Compare the affected user's role, accessible dashboards, enabled account state, authentication failures, and any conditional access or external identity policy. Test a harmless page the account should access before testing administrative functions. Do not grant administrator access merely to make the symptom disappear.
An app-specific access case showed the same device could use a browser but the companion app failed externally with an authentication-token request. The browser-versus-app authentication split demonstrates that one user-facing failure can occur after the server login page works.
If the affected account fails on every known-good client, repair or recreate only its authentication state according to the configured provider, then restore the minimum intended role. If another account fails on the same client too, identity is not the primary branch.
Reset Only the Affected Client State
Open a private browser profile or a different browser before clearing data. If the private profile works, remove Home Assistant site data or reset the companion frontend cache only on the affected client. Re-enter the exact known-good URL rather than relying on an old bookmark or autodiscovery result.
A version-scoped report of clients unable to connect documented extensive app reinstall, DNS, Wi-Fi, cellular, and URL tests. That scoped client connection failure supports preserving exact versions and comparisons without generalizing the case into a universal cause.
If a clean profile works, the repair is local client state; confirm login and live updates after closing and reopening the app. If every client profile fails only on that device, move to certificate, time, DNS, and network checks rather than resetting Home Assistant.
Compare URL, Certificate, DNS, and Network Path
Check device time, resolved address, certificate name and trust, selected internal or external URL, VPN state, and route on both Wi-Fi and cellular. A device can reach one URL through a browser while an app selects another stored URL or applies stricter certificate validation.
Use the ZimaSpace client-versus-server comparison to keep the healthy server path separate from one device's DNS, cache, or app behavior.
If the failure switches with the network, inspect split DNS, VLAN rules, captive portals, private relays, or VPN routes. If it follows one URL, correct that endpoint and certificate chain. Keep the working path unchanged while testing the failing one.
Confirm Recovery Without Widening the System
Repeat the original login or dashboard action on the affected account and device, then close and reopen the client, switch away and back to the original network, and verify live state updates. Also confirm a known-good user and device still work with no new privilege or proxy changes.
PASS means the scoped client remains connected and its intended permissions hold across those transitions. If the fix required administrator access, disabling certificate checks, or opening firewall rules broadly, roll it back; that is not a valid recovery.
Escalate when the exact failure reproduces on a clean client with a known-good account and correct network path, or when logs show a version-specific authentication error. Preserve versions, timestamps, URL class, and cross-test results so the report remains narrow and actionable.
Recommended products
Support & Tips
More to Read
Home Assistant Works on Wi-Fi but Fails on Ethernet or VPN
Test each network path separately, verify interface and routing state, distinguish direct IP from discovery, then repair only the failed layer.
How to Decommission Home Assistant Without Leaving Unprotected Data
Prove the replacement or archive, revoke every trust path, sanitize each data-bearing device, and retain only documented protected recovery copies.
Should You Use Automatic Updates for Home Assistant on a Home Server?
Choose manual, notification-only, or staged automatic updates from household impact, compatibility risk, observation time, and recovery readiness.
