fix(auth): prevent cookie loss during remote server timeout storms

Server timeouts caused orphaned fetchActivityData calls to fire clearCookieCache()
asynchronously, destroying cookies for all concurrent callers. Three fixes:

1. Replace Promise.race timeout with AbortController to properly cancel
   orphaned fetches and prevent delayed clearCookieCache() calls
2. Add cookie backup/restore — backupCookies() before clearCookieCache(),
   restoreCookieBackup() if re-login fails, so cookies are never lost
3. Add 15s auth failure throttle to block thundering herd re-logins when
   server slowdowns generate many 500 errors simultaneously

engage-api/get-activity.ts

@@ -1,12 +1,15 @@
 // engage-api/get-activity.ts
-import axios from 'axios';
+import axios, { type AxiosRequestConfig } from 'axios';
 import { logger } from '../utils/logger';
 import {
   ensureSingleLogin,
-  loadCachedCookies,
   saveCookiesToCache,
   clearCookieCache,
-  getCachedCookieString
+  getCachedCookieString,
+  backupCookies,
+  restoreCookieBackup,
+  tryAcquireAuthLock,
+  releaseAuthCooldown
 } from '../services/playwright-auth';
 
 // Define interfaces for our data structures
@@ -51,7 +54,8 @@ async function getActivityDetailsRaw(
   activityId: string,
   cookies: string,
   maxRetries: number = 3,
-  timeoutMilliseconds: number = 10000
+  timeoutMilliseconds: number = 10000,
+  signal?: AbortSignal
 ): Promise<string | null> {
   const url = 'https://engage.nkcswx.cn/Services/ActivitiesService.asmx/GetActivityDetails';
   const headers = {
@@ -65,13 +69,17 @@ async function getActivityDetailsRaw(
   };
 
   for (let attempt = 0; attempt < maxRetries; attempt++) {
+    if (signal?.aborted) {
+      logger.debug(`Activity ${activityId} aborted before attempt ${attempt + 1}`);
+      return null;
+    }
     try {
       logger.debug(`Attempt ${attempt + 1}/${maxRetries} for activity ${activityId} - Sending POST request to ${url}`);
       const response = await axios.post(url, payload, {
         headers,
         timeout: timeoutMilliseconds,
         responseType: 'text',
-        // Add additional timeout safety
+        signal,
         maxRedirects: 5
       });
       
@@ -155,6 +163,7 @@ export async function fetchActivityData(
   userName: string,
   userPwd: string,
   forceLogin: boolean = false,
+  signal?: AbortSignal
 ): Promise<any | null> {
   let currentCookie = forceLogin ? null : await getCachedCookieString();
 
@@ -164,17 +173,10 @@ export async function fetchActivityData(
     currentCookie = null;
   }
 
-  // Optimization: Skip pre-validation, directly request data
-  // Only validate/re-login when we get 4xx error OR after 5xx (backend may be in degraded state)
   if (!currentCookie) {
     logger.info('No cached cookie found. Attempting login...');
     try {
       currentCookie = await getCompleteCookies(userName, userPwd);
-      
-      const cookies = await loadCachedCookies();
-      if (cookies) {
-        await saveCookiesToCache(cookies);
-      }
     } catch (loginError) {
       logger.error(`Login process failed: ${(loginError as Error).message}`);
       return null;
@@ -187,37 +189,41 @@ export async function fetchActivityData(
   }
 
 
-
   logger.debug('Using cached cookie for API request.');
-  
+
   try {
     logger.debug(`Calling getActivityDetailsRaw for activity ${activityId}...`);
-    const rawActivityDetailsString = await getActivityDetailsRaw(activityId, currentCookie);
+    const rawActivityDetailsString = await getActivityDetailsRaw(activityId, currentCookie, 3, 10000, signal);
     logger.debug(`getActivityDetailsRaw returned for activity ${activityId}`);
     if (rawActivityDetailsString) {
       const parsedOuter = JSON.parse(rawActivityDetailsString);
       return JSON.parse(parsedOuter.d);
     }
-    // Check if this was a 5xx error and set flag for cookie validation
     logger.warn(`No data returned from getActivityDetailsRaw for activity ${activityId}, but no authentication error was thrown.`);
     return null;
   } catch (error) {
+    if (signal?.aborted) {
+      logger.debug(`Activity ${activityId} fetch aborted.`);
+      return null;
+    }
     if (error instanceof AuthenticationError) {
-      // Cookie returned 4xx, now validate and re-login
-      logger.warn(`API returned 4xx error (Status: ${error.status}). Cookie may be invalid. Attempting re-login and retry.`);
+      // Throttle: prevent thundering herd from multiple 500 errors
+      if (!tryAcquireAuthLock()) {
+        logger.info(`Auth throttled for activity ${activityId}. Reusing current cookies — likely still valid.`);
+        return null;
+      }
+
+      // Backup cookies before clearing so we can restore on re-login failure
+      backupCookies();
       await clearCookieCache();
 
       try {
         logger.info('Attempting re-login due to authentication failure...');
         currentCookie = await getCompleteCookies(userName, userPwd);
-        
-        const cookies = await loadCachedCookies();
-        if (cookies) {
-          await saveCookiesToCache(cookies);
-        }
+        releaseAuthCooldown();
 
         logger.info('Re-login successful. Retrying request for activity details...');
-        const rawActivityDetailsStringRetry = await getActivityDetailsRaw(activityId, currentCookie);
+        const rawActivityDetailsStringRetry = await getActivityDetailsRaw(activityId, currentCookie, 1, 10000, signal);
         if (rawActivityDetailsStringRetry) {
           const parsedOuterRetry = JSON.parse(rawActivityDetailsStringRetry);
           return JSON.parse(parsedOuterRetry.d);
@@ -225,7 +231,9 @@ export async function fetchActivityData(
         logger.warn(`Still no details for activity ${activityId} after re-login and retry.`);
         return null;
       } catch (retryLoginOrFetchError) {
-        logger.error(`Error during re-login or retry fetch for activity ${activityId}: ${(retryLoginOrFetchError as Error).message}`);
+        logger.error(`Re-login or retry failed for activity ${activityId}: ${(retryLoginOrFetchError as Error).message}`);
+        // Restore old cookies instead of leaving cache empty
+        await restoreCookieBackup();
         return null;
       }
     } else {

services/cache-manager.ts

@@ -52,14 +52,33 @@ async function processAndCacheActivity(activityId: string, forceUpdate: boolean
       throw new Error('API username or password not configured');
     }
     
-    // Add timeout protection for the entire fetch operation
+    // Add timeout protection via AbortController - properly cancels orphaned fetches
     logger.debug(`Fetching activity data for ID: ${activityId}`);
-    const activityJson = await Promise.race([
-      fetchActivityData(activityId, USERNAME, PASSWORD, false),
-      new Promise((_, reject) => 
-        setTimeout(() => reject(new Error(`Timeout fetching activity ${activityId} after ${CRAWLER_REQUEST_TIMEOUT_MS}ms`)), CRAWLER_REQUEST_TIMEOUT_MS + 5000)
-      )
-    ]);
+    const controller = new AbortController();
+    const timeoutId = setTimeout(
+      () => controller.abort(),
+      CRAWLER_REQUEST_TIMEOUT_MS + 5000
+    );
+
+    let activityJson: any = null;
+    try {
+      activityJson = await fetchActivityData(
+        activityId,
+        USERNAME,
+        PASSWORD,
+        false,
+        controller.signal
+      );
+    } finally {
+      clearTimeout(timeoutId);
+    }
+
+    if (controller.signal.aborted) {
+      logger.warn(`Request for activity ${activityId} timed out after ${CRAWLER_REQUEST_TIMEOUT_MS + 5000}ms. Cancelling orphaned fetch.`);
+      // Preserve existing cache on timeout
+      const existingData = await getActivityData(activityId);
+      return existingData || { lastCheck: new Date().toISOString(), error: `Timeout after ${CRAWLER_REQUEST_TIMEOUT_MS + 5000}ms` };
+    }
     let structuredActivity: ActivityData;
 
     if (!activityJson) {

services/playwright-auth.ts

@@ -11,6 +11,39 @@ let _inMemoryCookies: Cookie[] | null = null;
 // Login lock to prevent concurrent login attempts
 let _loginLock: Promise<Cookie[]> | null = null;
 
+// Cookie backup: preserved before clearCookieCache, restored on re-login failure
+let _cookieBackup: Cookie[] | null = null;
+
+// Auth failure throttle: debounce consecutive re-login triggers from 500 errors
+// Prevents thundering herd when server is slow and returns many 500s
+let _authFailureCooldownUntil = 0;
+const AUTH_FAILURE_COOLDOWN_MS = 15000; // 15s cooldown between re-login cycles
+
+/**
+ * Put all callers to wait during auth cooldown window.
+ * Returns true if auth is allowed (outside cooldown), false if throttled.
+ */
+export function tryAcquireAuthLock(): boolean {
+  const now = Date.now();
+  if (now < _authFailureCooldownUntil) {
+    const remaining = _authFailureCooldownUntil - now;
+    logger.warn(
+      `Re-login throttled: ${Math.round(remaining / 1000)}s cooldown remaining. ` +
+      `Existing cookies are likely still valid — server 500 is a temporary slowdown.`
+    );
+    return false;
+  }
+  return true;
+}
+
+/**
+ * Called after a successful re-login to release the cooldown.
+ */
+export function releaseAuthCooldown(): void {
+  _authFailureCooldownUntil = Date.now() + AUTH_FAILURE_COOLDOWN_MS;
+  logger.info(`Auth cooldown set: ${AUTH_FAILURE_COOLDOWN_MS}ms to prevent thundering herd re-logins.`);
+}
+
 /**
  * Ensure only one login process runs at a time
  */
@@ -178,8 +211,40 @@ export async function saveCookiesToCache(cookies: Cookie[]): Promise<void> {
   }
 }
 
+/**
+ * Backup current cookies before clearing. Restored if re-login fails.
+ */
+export function backupCookies(): Cookie[] | null {
+  if (_inMemoryCookies) {
+    _cookieBackup = [..._inMemoryCookies];
+    logger.info('Cookies backed up before clear.');
+  }
+  return _cookieBackup;
+}
+
+/**
+ * Restore cookies from backup after failed re-login.
+ */
+export async function restoreCookieBackup(): Promise<boolean> {
+  if (_cookieBackup) {
+    _inMemoryCookies = _cookieBackup;
+    try {
+      await fs.promises.writeFile(COOKIE_FILE_PATH, JSON.stringify(_cookieBackup, null, 2), 'utf-8');
+      logger.info('Cookies restored from backup successfully.');
+      _cookieBackup = null;
+      return true;
+    } catch (error: any) {
+      logger.error('Failed to restore cookies from backup:', error.message);
+      return false;
+    }
+  }
+  logger.warn('No cookie backup available for restore.');
+  return false;
+}
+
 /**
  * Clear cookie cache
+ * Prefer backupAndClearCookieCache() instead to preserve old cookies.
  */
 export async function clearCookieCache(): Promise<void> {
   _inMemoryCookies = null;