fix(update): authenticate restart health probes

vincentkoc · vincentkoc · commit b546aa91e19b · 2026-05-04T20:14:02.000-07:00
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -72,6 +72,7 @@ Docs: https://docs.openclaw.ai
 - Telegram/media: derive no-caption inbound media placeholders from saved MIME metadata instead of the Telegram `photo` shape, so non-image and mixed attachments no longer reach the model as `<media:image>`. Fixes #69793. Thanks @aspalagin.
 - Agents/cache: keep per-turn runtime context out of ordinary chat system prompts while still delivering hidden current-turn context, restoring prompt-cache reuse on chat continuations. Fixes #77431. Thanks @Udjin79.
 - Gateway/startup: include resolved thinking and fast-mode defaults in the `agent model` startup log line, defaulting unset startup thinking to `medium` without mixing in reasoning visibility.
+- Gateway/update: resolve local gateway probe auth from the installed config during post-update restart verification, so token/device-authenticated VPS gateways are not misreported as unhealthy port conflicts after a package swap. Thanks @vincentkoc.
 - Agents/Tools: add post-compaction loop guard in `pi-embedded-runner` that arms after auto-compaction-retry and aborts the run with `compaction_loop_persisted` when the agent emits the same `(tool, args, result)` triple `windowSize` times (default 3) within that window. Disable via existing `tools.loopDetection.enabled`; tune via `tools.loopDetection.postCompactionGuard.windowSize`. Targets the failure mode where context-overflow + compaction does not break a tool-call loop. Refs #77474; carries forward #21597. Thanks @efpiva.
 - Gateway/watch: suppress sync-I/O trace output during `pnpm gateway:watch --benchmark` unless explicitly requested, so CPU profiling no longer floods the terminal with stack traces.
 - Gateway/watch: when benchmark sync-I/O tracing is explicitly enabled, tee trace blocks to the benchmark output log and filter them from the terminal pane while keeping normal Gateway logs visible.
diff --git a/src/cli/daemon-cli/restart-health.test.ts b/src/cli/daemon-cli/restart-health.test.ts
@@ -8,6 +8,12 @@ const classifyPortListener = vi.hoisted(() =>
   vi.fn<(_listener: unknown, _port: number) => PortListenerKind>(() => "gateway"),
 );
 const probeGateway = vi.hoisted(() => vi.fn());
+const readBestEffortConfig = vi.hoisted(() => vi.fn(async () => ({})));
+const resolveGatewayProbeAuthSafeWithSecretInputs = vi.hoisted(() =>
+  vi.fn<(_opts: unknown) => Promise<{ auth: { token?: string; password?: string } }>>(async () => ({
+    auth: {},
+  })),
+);
 
 vi.mock("../../infra/ports.js", () => ({
   classifyPortListener: (listener: unknown, port: number) => classifyPortListener(listener, port),
@@ -19,6 +25,17 @@ vi.mock("../../gateway/probe.js", () => ({
   probeGateway: (opts: unknown) => probeGateway(opts),
 }));
 
+vi.mock("../../config/io.js", () => ({
+  createConfigIO: () => ({
+    readBestEffortConfig: () => readBestEffortConfig(),
+  }),
+}));
+
+vi.mock("../../gateway/probe-auth.js", () => ({
+  resolveGatewayProbeAuthSafeWithSecretInputs: (opts: unknown) =>
+    resolveGatewayProbeAuthSafeWithSecretInputs(opts),
+}));
+
 vi.mock("../../utils.js", async () => {
   const actual = await vi.importActual<typeof import("../../utils.js")>("../../utils.js");
   return {
@@ -112,6 +129,10 @@ async function waitForStoppedFreeGatewayRestart() {
 describe("inspectGatewayRestart", () => {
   beforeEach(() => {
     inspectPortUsage.mockReset();
+    readBestEffortConfig.mockReset();
+    readBestEffortConfig.mockResolvedValue({});
+    resolveGatewayProbeAuthSafeWithSecretInputs.mockReset();
+    resolveGatewayProbeAuthSafeWithSecretInputs.mockResolvedValue({ auth: {} });
     inspectPortUsage.mockResolvedValue({
       port: 0,
       status: "free",
@@ -380,6 +401,52 @@ describe("inspectGatewayRestart", () => {
     expect(snapshot.versionMismatch).toBeUndefined();
   });
 
+  it("uses configured local probe auth while waiting for a matching-version restart", async () => {
+    readBestEffortConfig.mockResolvedValue({
+      gateway: { auth: { mode: "token", token: "probe-token" } },
+    });
+    resolveGatewayProbeAuthSafeWithSecretInputs.mockResolvedValue({
+      auth: { token: "probe-token" },
+    });
+    probeGateway.mockResolvedValue({
+      ok: true,
+      close: null,
+      server: { version: "2026.4.24", connId: "new" },
+    });
+    const service = makeGatewayService({ status: "running", pid: 8000 });
+    inspectPortUsage.mockResolvedValue({
+      port: 18789,
+      status: "busy",
+      listeners: [{ pid: 8000, commandLine: "openclaw-gateway" }],
+      hints: [],
+    });
+
+    const { waitForGatewayHealthyRestart } = await import("./restart-health.js");
+    const snapshot = await waitForGatewayHealthyRestart({
+      service,
+      port: 18789,
+      expectedVersion: "2026.4.24",
+      attempts: 1,
+    });
+
+    expect(snapshot).toMatchObject({
+      healthy: true,
+      gatewayVersion: "2026.4.24",
+      expectedVersion: "2026.4.24",
+    });
+    expect(resolveGatewayProbeAuthSafeWithSecretInputs).toHaveBeenCalledWith(
+      expect.objectContaining({
+        cfg: { gateway: { auth: { mode: "token", token: "probe-token" } } },
+        mode: "local",
+      }),
+    );
+    expect(probeGateway).toHaveBeenCalledWith(
+      expect.objectContaining({
+        auth: { token: "probe-token", password: undefined },
+      }),
+    );
+  });
+
   it("stops waiting once the restarted gateway reports the wrong version", async () => {
     probeGateway.mockResolvedValue({
       ok: true,
diff --git a/src/cli/daemon-cli/restart-health.ts b/src/cli/daemon-cli/restart-health.ts
@@ -1,6 +1,9 @@
 import type { PluginHealthErrorSummary } from "../../commands/health.types.js";
+import { createConfigIO } from "../../config/io.js";
+import type { OpenClawConfig } from "../../config/types.openclaw.js";
 import type { GatewayServiceRuntime } from "../../daemon/service-runtime.js";
 import type { GatewayService } from "../../daemon/service.js";
+import { resolveGatewayProbeAuthSafeWithSecretInputs } from "../../gateway/probe-auth.js";
 import { probeGateway } from "../../gateway/probe.js";
 import {
   classifyPortListener,
@@ -61,6 +64,11 @@ type GatewayReachability = {
   channelProbeErrors: Array<{ id: string; error: string }>;
 };
 
+type GatewayRestartProbeAuth = {
+  token?: string;
+  password?: string;
+};
+
 function hasListenerAttributionGap(portUsage: PortUsage): boolean {
   if (portUsage.status !== "busy" || portUsage.listeners.length > 0) {
     return false;
@@ -228,9 +236,12 @@ function applyChannelProbeErrors(snapshot: GatewayRestartSnapshot): GatewayResta
 async function confirmGatewayReachable(params: {
   port: number;
   includeHealthDetails?: boolean;
+  auth?: GatewayRestartProbeAuth;
 }): Promise<GatewayReachability> {
-  const token = normalizeOptionalString(process.env.OPENCLAW_GATEWAY_TOKEN);
-  const password = normalizeOptionalString(process.env.OPENCLAW_GATEWAY_PASSWORD);
+  const token = normalizeOptionalString(params.auth?.token ?? process.env.OPENCLAW_GATEWAY_TOKEN);
+  const password = normalizeOptionalString(
+    params.auth?.password ?? process.env.OPENCLAW_GATEWAY_PASSWORD,
+  );
   const probe = await probeGateway({
     url: `ws://127.0.0.1:${params.port}`,
     auth: token || password ? { token, password } : undefined,
@@ -251,13 +262,37 @@ async function confirmGatewayReachable(params: {
   };
 }
 
-async function inspectGatewayPortHealth(port: number): Promise<GatewayPortHealthSnapshot> {
+async function resolveGatewayRestartProbeAuth(
+  env: NodeJS.ProcessEnv | undefined,
+): Promise<GatewayRestartProbeAuth | undefined> {
+  const mergedEnv = {
+    ...(process.env as Record<string, string | undefined>),
+    ...(env ?? undefined),
+  } as NodeJS.ProcessEnv;
+  const cfg = await createConfigIO({
+    env: mergedEnv,
+    pluginValidation: "skip",
+  })
+    .readBestEffortConfig()
+    .catch((): OpenClawConfig => ({}));
+  const resolved = await resolveGatewayProbeAuthSafeWithSecretInputs({
+    cfg,
+    mode: "local",
+    env: mergedEnv,
+  });
+  return resolved.auth;
+}
+
+async function inspectGatewayPortHealth(params: {
+  port: number;
+  auth?: GatewayRestartProbeAuth;
+}): Promise<GatewayPortHealthSnapshot> {
   let portUsage: PortUsage;
   try {
-    portUsage = await inspectPortUsage(port);
+    portUsage = await inspectPortUsage(params.port);
   } catch (err) {
     portUsage = {
-      port,
+      port: params.port,
       status: "unknown",
       listeners: [],
       hints: [],
@@ -268,7 +303,12 @@ async function inspectGatewayPortHealth(port: number): Promise<GatewayPortHealth
   let healthy = false;
   if (portUsage.status === "busy") {
     try {
-      healthy = (await confirmGatewayReachable({ port })).reachable;
+      healthy = (
+        await confirmGatewayReachable({
+          port: params.port,
+          auth: params.auth,
+        })
+      ).reachable;
     } catch {
       // best-effort probe
     }
@@ -283,6 +323,7 @@ export async function inspectGatewayRestart(params: {
   env?: NodeJS.ProcessEnv;
   expectedVersion?: string | null;
   includeUnknownListenersAsStale?: boolean;
+  probeAuth?: GatewayRestartProbeAuth;
 }): Promise<GatewayRestartSnapshot> {
   const env = params.env ?? process.env;
   const expectedVersion = normalizeOptionalString(params.expectedVersion);
@@ -294,6 +335,7 @@ export async function inspectGatewayRestart(params: {
       reachability = await confirmGatewayReachable({
         port: params.port,
         includeHealthDetails: Boolean(expectedVersion),
+        auth: params.probeAuth,
       });
       activatedPluginErrors = reachability.activatedPluginErrors;
       channelProbeErrors = reachability.channelProbeErrors;
@@ -477,12 +519,14 @@ export async function waitForGatewayHealthyRestart(params: {
   const attempts = params.attempts ?? DEFAULT_RESTART_HEALTH_ATTEMPTS;
   const delayMs = params.delayMs ?? DEFAULT_RESTART_HEALTH_DELAY_MS;
 
+  const probeAuth = await resolveGatewayRestartProbeAuth(params.env).catch(() => undefined);
   let snapshot = await inspectGatewayRestart({
     service: params.service,
     port: params.port,
     env: params.env,
     expectedVersion: params.expectedVersion,
     includeUnknownListenersAsStale: params.includeUnknownListenersAsStale,
+    probeAuth,
   });
 
   let consecutiveStoppedFreeCount = 0;
@@ -523,6 +567,7 @@ export async function waitForGatewayHealthyRestart(params: {
       env: params.env,
       expectedVersion: params.expectedVersion,
       includeUnknownListenersAsStale: params.includeUnknownListenersAsStale,
+      probeAuth,
     });
   }
 
@@ -537,14 +582,21 @@ export async function waitForGatewayHealthyListener(params: {
   const attempts = params.attempts ?? DEFAULT_RESTART_HEALTH_ATTEMPTS;
   const delayMs = params.delayMs ?? DEFAULT_RESTART_HEALTH_DELAY_MS;
 
-  let snapshot = await inspectGatewayPortHealth(params.port);
+  const probeAuth = await resolveGatewayRestartProbeAuth(undefined).catch(() => undefined);
+  let snapshot = await inspectGatewayPortHealth({
+    port: params.port,
+    auth: probeAuth,
+  });
 
   for (let attempt = 0; attempt < attempts; attempt += 1) {
     if (snapshot.healthy) {
       return snapshot;
     }
     await sleep(delayMs);
-    snapshot = await inspectGatewayPortHealth(params.port);
+    snapshot = await inspectGatewayPortHealth({
+      port: params.port,
+      auth: probeAuth,
+    });
   }
 
   return snapshot;
