fix(gateway): drain active turns before restart to prevent message loss (#13931)

* fix(gateway): drain active turns before restart to prevent message loss

On SIGUSR1 restart, the gateway now waits up to 30s for in-flight agent
turns to complete before tearing down the server. This prevents buffered
messages from being dropped when config.patch or update triggers a restart
while agents are mid-turn.

Changes:
- command-queue.ts: add getActiveTaskCount() and waitForActiveTasks()
  helpers to track and wait on active lane tasks
- run-loop.ts: on restart signal, drain active tasks before server.close()
  with a 30s timeout; extend force-exit timer accordingly
- command-queue.test.ts: update imports for new exports

Fixes #13883

* fix(queue): snapshot active tasks for restart drain

---------

Co-authored-by: Elonito <0xRaini@users.noreply.github.com>
Co-authored-by: Tak Hoffman <781889+Takhoffman@users.noreply.github.com>

src/cli/gateway-cli/run-loop.ts

@@ -6,6 +6,7 @@ import {
   isGatewaySigusr1RestartExternallyAllowed,
 } from "../../infra/restart.js";
 import { createSubsystemLogger } from "../../logging/subsystem.js";
+import { getActiveTaskCount, waitForActiveTasks } from "../../process/command-queue.js";
 
 const gatewayLog = createSubsystemLogger("gateway");
 
@@ -26,6 +27,9 @@ export async function runGatewayLoop(params: {
     process.removeListener("SIGUSR1", onSigusr1);
   };
 
+  const DRAIN_TIMEOUT_MS = 30_000;
+  const SHUTDOWN_TIMEOUT_MS = 5_000;
+
   const request = (action: GatewayRunSignalAction, signal: string) => {
     if (shuttingDown) {
       gatewayLog.info(`received ${signal} during shutdown; ignoring`);
@@ -35,14 +39,33 @@ export async function runGatewayLoop(params: {
     const isRestart = action === "restart";
     gatewayLog.info(`received ${signal}; ${isRestart ? "restarting" : "shutting down"}`);
 
+    // Allow extra time for draining active turns on restart.
+    const forceExitMs = isRestart ? DRAIN_TIMEOUT_MS + SHUTDOWN_TIMEOUT_MS : SHUTDOWN_TIMEOUT_MS;
     const forceExitTimer = setTimeout(() => {
       gatewayLog.error("shutdown timed out; exiting without full cleanup");
       cleanupSignals();
       params.runtime.exit(0);
-    }, 5000);
+    }, forceExitMs);
 
     void (async () => {
       try {
+        // On restart, wait for in-flight agent turns to finish before
+        // tearing down the server so buffered messages are delivered.
+        if (isRestart) {
+          const activeTasks = getActiveTaskCount();
+          if (activeTasks > 0) {
+            gatewayLog.info(
+              `draining ${activeTasks} active task(s) before restart (timeout ${DRAIN_TIMEOUT_MS}ms)`,
+            );
+            const { drained } = await waitForActiveTasks(DRAIN_TIMEOUT_MS);
+            if (drained) {
+              gatewayLog.info("all active tasks drained");
+            } else {
+              gatewayLog.warn("drain timeout reached; proceeding with restart");
+            }
+          }
+        }
+
         await server?.close({
           reason: isRestart ? "gateway restarting" : "gateway stopping",
           restartExpectedMs: isRestart ? 1500 : null,