fix(gateway): health check timeout when lsof is not installed

riftzen-bit · riftzen-bit · commit 45fc88cd9f93 · 2026-03-02T21:49:38.000-08:00
The `ownsPort` check in `inspectGatewayRestart` only verified port
ownership via `portUsage.listeners.some(...)` when `runtimePid` was
known.  When `lsof` is not installed, `inspectPortUsage` returns
`{ status: "busy", listeners: [] }` because `checkPortInUse` detects
the port is occupied but cannot enumerate listeners.  `.some()` on an
empty array always returns `false`, so `ownsPort` was always `false`,
causing the health-check loop to spin for the full 60 s timeout on
every restart.

Add the same `(status === "busy" &amp;&amp; listeners.length === 0)` fallback
that already existed in the `runtimePid == null` branch so that a
running service with a known PID is treated as the port owner when
listener enumeration is unavailable.
diff --git a/src/cli/daemon-cli/restart-health.test.ts b/src/cli/daemon-cli/restart-health.test.ts
@@ -123,6 +123,26 @@ describe("inspectGatewayRestart", () => {
     expect(snapshot.staleGatewayPids).toEqual([]);
   });
 
+  it("treats port as owned when runtime pid is known but listeners are empty (e.g. lsof missing)", async () => {
+    const service = {
+      readRuntime: vi.fn(async () => ({ status: "running", pid: 40977 })),
+    } as unknown as GatewayService;
+
+    inspectPortUsage.mockResolvedValue({
+      port: 18789,
+      status: "busy",
+      listeners: [],
+      hints: [],
+      errors: ["Error: spawn lsof ENOENT"],
+    });
+
+    const { inspectGatewayRestart } = await import("./restart-health.js");
+    const snapshot = await inspectGatewayRestart({ service, port: 18789 });
+
+    expect(snapshot.healthy).toBe(true);
+    expect(snapshot.staleGatewayPids).toEqual([]);
+  });
+
   it("does not treat known non-gateway listeners as stale in fallback mode", async () => {
     Object.defineProperty(process, "platform", { value: "win32", configurable: true });
     classifyPortListener.mockReturnValue("ssh");
diff --git a/src/cli/daemon-cli/restart-health.ts b/src/cli/daemon-cli/restart-health.ts
@@ -74,9 +74,15 @@ export async function inspectGatewayRestart(params: {
       : [];
   const running = runtime.status === "running";
   const runtimePid = runtime.pid;
+  // When the port is busy but no listeners are available (e.g. lsof not installed),
+  // optimistically assume the running service owns the port since we cannot
+  // distinguish ownership without process enumeration tools.
   const ownsPort =
     runtimePid != null
-      ? portUsage.listeners.some((listener) => listenerOwnedByRuntimePid({ listener, runtimePid }))
+      ? portUsage.listeners.some((listener) =>
+          listenerOwnedByRuntimePid({ listener, runtimePid }),
+        ) ||
+        (portUsage.status === "busy" && portUsage.listeners.length === 0)
       : gatewayListeners.length > 0 ||
         (portUsage.status === "busy" && portUsage.listeners.length === 0);
   const healthy = running && ownsPort;
