Recover stale daemon connections with liveness probes (#1124)

Use top-level WebSocket ping/pong for client liveness so session RPC timeouts stay scoped to the operation that timed out. Back off inactive connection probes while a host already has a healthy active connection.
2026-05-21 11:43:22 +08:00
parent 28415b3542
commit dee8f485de
5 changed files with 285 additions and 2 deletions
@@ -146,6 +146,8 @@ There is no dedicated welcome message; the server emits a `status` session messa

 **Top-level WS envelopes** are `hello`, `recording_state`, `ping`/`pong`, and `session` (which wraps the rich union of session messages).

+Client liveness checks use the top-level JSON `ping`/`pong` envelope, not a session RPC and not RFC6455 protocol ping. The app runs through browser and React Native WebSocket APIs, which do not expose protocol ping, so this envelope is the portable way to test the direct or relay data path. Session RPC timeouts are operation failures and must not be treated as proof that the socket is dead.
+
 New session RPCs use dotted names with `.request` and `.response` suffixes, such as `checkout.github.set_auto_merge.request` and `checkout.github.set_auto_merge.response`. See [rpc-namespacing.md](rpc-namespacing.md) for the convention and migration rules for older flat RPC names.

 **Notable session message types:**