Route policy and diagnosis
A device can have Direct, SSH and Cloud routes. Automatic selection reduces manual switching, but diagnose each route independently first.
Inspect and test
anytty endpoint show workstation
anytty endpoint policy show workstation
anytty endpoint test workstation --route ROUTE_ID --jsonUse real route IDs and record each result. Listing endpoints reads saved configuration; test attempts protocol reachability.
Configure the next connection
anytty endpoint policy set workstation --route autoRoute policy accepts auto, direct, ssh or cloud. Cloud paths support auto, p2p, relay or smart_route; Relay transport supports auto, udp or tcp. Available parameters do not guarantee service-plan entitlement.
Policies apply to the next session. Reconnect and inspect the actual route. Forced routes help diagnosis; verify backups before restoring automatic selection.
Diagnose by layer
| Symptom | Check first |
|---|---|
| Timeout/refused connection | Target, listener, network and firewall |
| SSH authentication failure | Account, keys and host verification |
| Daemon identity mismatch | Wrong device, reinstall or identity change |
| Authorization failure | Redemption, expiration or revocation |
| Connected but task missing | Endpoint, filters and terminal ID |
Fallback changes routing, not identity or authorization requirements. Reports should include client/daemon versions, time, route type and redacted errors, never complete invitations or credentials.