From e0f432bc39b559ea364f2adfc60355e22ec0f944 Mon Sep 17 00:00:00 2001
From: "Gan, Jimmy" <jimmy.gan@adidas.com>
Date: Wed, 11 Mar 2026 23:55:51 +0800
Subject: [PATCH] fix: add pong/timeout detection to prevent silent connection
 hangs

- Backend now responds to __ping__ with __pong__
- Frontend detects missing pong within 5s and closes connection
- This triggers proper reconnect/auth recovery instead of hanging
- Fixes 'keepalive ping timeout' issue where connection appears alive but is dead
---
 dashboard/backend/routers/terminal.py         |  1 +
 dashboard/frontend/src/routes/Terminal.svelte | 26 ++++++++++++++++++-
 2 files changed, 26 insertions(+), 1 deletion(-)

diff --git a/dashboard/backend/routers/terminal.py b/dashboard/backend/routers/terminal.py
index 8750a02..90ea7b6 100644
--- a/dashboard/backend/routers/terminal.py
+++ b/dashboard/backend/routers/terminal.py
@@ -180,6 +180,7 @@ async def ws_endpoint(websocket: WebSocket, host: str = Query("nas"), token: str
                         process.stdin.write(data)
                 elif "text" in msg and msg["text"]:
                     if msg["text"] == "__ping__":
+                        await websocket.send_text("__pong__")
                         continue
                     process.stdin.write(msg["text"].encode())
         except Exception as e:
diff --git a/dashboard/frontend/src/routes/Terminal.svelte b/dashboard/frontend/src/routes/Terminal.svelte
index 273147c..d297558 100644
--- a/dashboard/frontend/src/routes/Terminal.svelte
+++ b/dashboard/frontend/src/routes/Terminal.svelte
@@ -190,18 +190,24 @@
 
       const proto = location.protocol === "https:" ? "wss:" : "ws:";
       let heartbeat = null;
+      let pongTimeout = null;
       let reconnectTimer = null;
       let closedManually = false;
       let reconnecting = false;
       let authRecoveryInFlight = false;
       let reconnectAttempts = 0;
       const MAX_RECONNECT_DELAY = 30000;
+      const PONG_TIMEOUT = 5000; // 5s to receive pong
 
       function clearHeartbeat() {
         if (heartbeat) {
           clearInterval(heartbeat);
           heartbeat = null;
         }
+        if (pongTimeout) {
+          clearTimeout(pongTimeout);
+          pongTimeout = null;
+        }
       }
 
       function handleAuthExpired(reason = "Session expired — please log in again") {
@@ -280,7 +286,15 @@
           sendSize(ws);
           clearHeartbeat();
           heartbeat = setInterval(() => {
-            if (ws.readyState === 1) ws.send("__ping__");
+            if (ws.readyState === 1) {
+              ws.send("__ping__");
+              // Set timeout to detect missing pong
+              pongTimeout = setTimeout(() => {
+                if (ws.readyState === 1) {
+                  ws.close(1000, "keepalive ping timeout");
+                }
+              }, PONG_TIMEOUT);
+            }
           }, 12000);
           const d = tabData.get(tabId);
           if (d) {
@@ -293,6 +307,16 @@
           }
         };
         ws.onmessage = (e) => {
+          // Check for text pong response
+          if (typeof e.data === 'string' && e.data === '__pong__') {
+            // Clear the pong timeout - connection is alive
+            if (pongTimeout) {
+              clearTimeout(pongTimeout);
+              pongTimeout = null;
+            }
+            return;
+          }
+
           const data = new Uint8Array(e.data);
           const isPersistenceMessage = tab.persistent
             && data.length >= PERSISTENCE_STATUS_PREFIX_BYTES.length