@ceon-oy/monitor-sdk 1.5.3 → 1.5.7

package/dist/index.d.mts

@@ -189,6 +189,18 @@ interface SdkHealthResultsResponse {
         total: number;
     };
 }
+interface ProjectSettings {
+    name: string;
+    vulnerabilityScanIntervalHours: number;
+    scanRequestedAt: string | null;
+    techScanRequestedAt: string | null;
+    metricsEnabled?: boolean;
+    metricsIntervalSeconds?: number;
+    metricsDiskPaths?: string[];
+    metricsCpuThreshold?: number;
+    metricsRamThreshold?: number;
+    metricsDiskThreshold?: number;
+}
 
 declare class MonitorClient {
     private apiKey;
@@ -291,18 +303,16 @@ declare class MonitorClient {
      * Fetch project settings from the monitoring server.
      * Returns configuration including vulnerability scan interval and scan request timestamps.
      */
-    fetchProjectSettings(): Promise<{
-        name: string;
-        vulnerabilityScanIntervalHours: number;
-        scanRequestedAt: string | null;
-        techScanRequestedAt: string | null;
-        metricsEnabled?: boolean;
-        metricsIntervalSeconds?: number;
-        metricsDiskPaths?: string[];
-        metricsCpuThreshold?: number;
-        metricsRamThreshold?: number;
-        metricsDiskThreshold?: number;
-    } | null>;
+    fetchProjectSettings(): Promise<ProjectSettings | null>;
+    /**
+     * Fetch project settings, retrying a couple of times on failure before giving up.
+     * A single transient failure here (a slow DNS lookup, a brief server hiccup at
+     * boot) used to permanently disable auto-audit for the process's entire
+     * lifetime, since this call only ever ran once at startup. Retrying a few
+     * seconds apart absorbs that kind of one-off blip instead of needing a full
+     * process restart to recover.
+     */
+    private fetchProjectSettingsWithRetry;
     /**
      * Setup automatic vulnerability scanning based on server-configured interval.
      * Fetches settings from server and sets up recurring scans.

package/dist/index.d.ts

@@ -189,6 +189,18 @@ interface SdkHealthResultsResponse {
         total: number;
     };
 }
+interface ProjectSettings {
+    name: string;
+    vulnerabilityScanIntervalHours: number;
+    scanRequestedAt: string | null;
+    techScanRequestedAt: string | null;
+    metricsEnabled?: boolean;
+    metricsIntervalSeconds?: number;
+    metricsDiskPaths?: string[];
+    metricsCpuThreshold?: number;
+    metricsRamThreshold?: number;
+    metricsDiskThreshold?: number;
+}
 
 declare class MonitorClient {
     private apiKey;
@@ -291,18 +303,16 @@ declare class MonitorClient {
      * Fetch project settings from the monitoring server.
      * Returns configuration including vulnerability scan interval and scan request timestamps.
      */
-    fetchProjectSettings(): Promise<{
-        name: string;
-        vulnerabilityScanIntervalHours: number;
-        scanRequestedAt: string | null;
-        techScanRequestedAt: string | null;
-        metricsEnabled?: boolean;
-        metricsIntervalSeconds?: number;
-        metricsDiskPaths?: string[];
-        metricsCpuThreshold?: number;
-        metricsRamThreshold?: number;
-        metricsDiskThreshold?: number;
-    } | null>;
+    fetchProjectSettings(): Promise<ProjectSettings | null>;
+    /**
+     * Fetch project settings, retrying a couple of times on failure before giving up.
+     * A single transient failure here (a slow DNS lookup, a brief server hiccup at
+     * boot) used to permanently disable auto-audit for the process's entire
+     * lifetime, since this call only ever ran once at startup. Retrying a few
+     * seconds apart absorbs that kind of one-off blip instead of needing a full
+     * process restart to recover.
+     */
+    private fetchProjectSettingsWithRetry;
     /**
      * Setup automatic vulnerability scanning based on server-configured interval.
      * Fetches settings from server and sets up recurring scans.

package/dist/index.js

@@ -53,10 +53,15 @@ var CONFIG_LIMITS = {
   // 10MB
   AUDIT_TIMEOUT_MS: 6e4,
   // 60 seconds (default)
-  MAX_AUDIT_TIMEOUT_MS: 3e5,
-  // 5 minutes (max configurable)
+  MAX_AUDIT_TIMEOUT_MS: 18e5,
+  // 30 minutes (max configurable) - was 5 minutes, which every
+  // real npm audit run on Cloud Run (slow registry egress) exceeded, silently returning null
   SETTINGS_POLL_INTERVAL_MS: 5 * 60 * 1e3,
   // 5 minutes
+  SETUP_SETTINGS_MAX_ATTEMPTS: 3,
+  // retry a transient failure fetching settings at startup
+  SETUP_SETTINGS_RETRY_DELAY_MS: 5e3,
+  // before falling back to the periodic poll to self-heal
   REGISTRY_TIMEOUT_MS: 5e3,
   // 5 seconds (default)
   MAX_REGISTRY_TIMEOUT_MS: 3e4,
@@ -489,15 +494,46 @@ var MonitorClient = class {
       return null;
     }
   }
+  /**
+   * Fetch project settings, retrying a couple of times on failure before giving up.
+   * A single transient failure here (a slow DNS lookup, a brief server hiccup at
+   * boot) used to permanently disable auto-audit for the process's entire
+   * lifetime, since this call only ever ran once at startup. Retrying a few
+   * seconds apart absorbs that kind of one-off blip instead of needing a full
+   * process restart to recover.
+   */
+  async fetchProjectSettingsWithRetry() {
+    for (let attempt = 1; attempt <= CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS; attempt++) {
+      const settings = await this.fetchProjectSettings();
+      if (settings) {
+        return settings;
+      }
+      if (attempt < CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS) {
+        console.warn(
+          `[MonitorClient] Fetching project settings failed (attempt ${attempt}/${CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS}), retrying...`
+        );
+        await new Promise((resolve) => setTimeout(resolve, CONFIG_LIMITS.SETUP_SETTINGS_RETRY_DELAY_MS));
+      }
+    }
+    return null;
+  }
   /**
    * Setup automatic vulnerability scanning based on server-configured interval.
    * Fetches settings from server and sets up recurring scans.
    * Also sets up polling for on-demand scan requests from the server.
    */
   async setupAutoAudit() {
-    const settings = await this.fetchProjectSettings();
+    const settings = await this.fetchProjectSettingsWithRetry();
+    console.log("[MonitorClient] Polling for scan requests enabled (every 5 minutes)");
+    this.settingsPollingTimer = setInterval(() => {
+      this.checkForScanRequest().catch((err) => {
+        this.reportError("SETTINGS_POLL", "Scan request check failed", err);
+      });
+    }, CONFIG_LIMITS.SETTINGS_POLL_INTERVAL_MS);
     if (!settings) {
-      console.warn("[MonitorClient] Could not fetch project settings, auto audit disabled");
+      const message = "Could not fetch project settings after retrying at startup; will keep retrying every 5 minutes";
+      console.warn(`[MonitorClient] ${message}`);
+      this.reportError("AUDIT_SETUP", message);
       return;
     }
     if (settings.scanRequestedAt) {
@@ -515,12 +551,6 @@ var MonitorClient = class {
       await this.runScanAndTrackTime();
       this.setAuditInterval(intervalHours);
     }
-    console.log("[MonitorClient] Polling for scan requests enabled (every 5 minutes)");
-    this.settingsPollingTimer = setInterval(() => {
-      this.checkForScanRequest().catch((err) => {
-        this.reportError("SETTINGS_POLL", "Scan request check failed", err);
-      });
-    }, CONFIG_LIMITS.SETTINGS_POLL_INTERVAL_MS);
   }
   /**
    * (Re)configure the recurring audit timer for the given interval, replacing
@@ -1261,7 +1291,9 @@ var MonitorClient = class {
           maxBuffer: CONFIG_LIMITS.AUDIT_MAX_BUFFER
         });
         if (result2.timedOut) {
-          console.error(`[MonitorClient] yarn audit timed out after ${this.auditTimeoutMs}ms`);
+          const message = `yarn audit timed out after ${this.auditTimeoutMs}ms`;
+          console.error(`[MonitorClient] ${message}`);
+          this.reportError("VULNERABILITY_SCAN", message);
           return null;
         }
         auditOutput = result2.stdout;
@@ -1294,7 +1326,9 @@ var MonitorClient = class {
           maxBuffer: CONFIG_LIMITS.AUDIT_MAX_BUFFER
         });
         if (result2.timedOut) {
-          console.error(`[MonitorClient] npm audit timed out after ${this.auditTimeoutMs}ms`);
+          const message = `npm audit timed out after ${this.auditTimeoutMs}ms`;
+          console.error(`[MonitorClient] ${message}`);
+          this.reportError("VULNERABILITY_SCAN", message);
           return null;
         }
         auditOutput = result2.stdout;

package/dist/index.mjs

@@ -17,10 +17,15 @@ var CONFIG_LIMITS = {
   // 10MB
   AUDIT_TIMEOUT_MS: 6e4,
   // 60 seconds (default)
-  MAX_AUDIT_TIMEOUT_MS: 3e5,
-  // 5 minutes (max configurable)
+  MAX_AUDIT_TIMEOUT_MS: 18e5,
+  // 30 minutes (max configurable) - was 5 minutes, which every
+  // real npm audit run on Cloud Run (slow registry egress) exceeded, silently returning null
   SETTINGS_POLL_INTERVAL_MS: 5 * 60 * 1e3,
   // 5 minutes
+  SETUP_SETTINGS_MAX_ATTEMPTS: 3,
+  // retry a transient failure fetching settings at startup
+  SETUP_SETTINGS_RETRY_DELAY_MS: 5e3,
+  // before falling back to the periodic poll to self-heal
   REGISTRY_TIMEOUT_MS: 5e3,
   // 5 seconds (default)
   MAX_REGISTRY_TIMEOUT_MS: 3e4,
@@ -453,15 +458,46 @@ var MonitorClient = class {
       return null;
     }
   }
+  /**
+   * Fetch project settings, retrying a couple of times on failure before giving up.
+   * A single transient failure here (a slow DNS lookup, a brief server hiccup at
+   * boot) used to permanently disable auto-audit for the process's entire
+   * lifetime, since this call only ever ran once at startup. Retrying a few
+   * seconds apart absorbs that kind of one-off blip instead of needing a full
+   * process restart to recover.
+   */
+  async fetchProjectSettingsWithRetry() {
+    for (let attempt = 1; attempt <= CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS; attempt++) {
+      const settings = await this.fetchProjectSettings();
+      if (settings) {
+        return settings;
+      }
+      if (attempt < CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS) {
+        console.warn(
+          `[MonitorClient] Fetching project settings failed (attempt ${attempt}/${CONFIG_LIMITS.SETUP_SETTINGS_MAX_ATTEMPTS}), retrying...`
+        );
+        await new Promise((resolve) => setTimeout(resolve, CONFIG_LIMITS.SETUP_SETTINGS_RETRY_DELAY_MS));
+      }
+    }
+    return null;
+  }
   /**
    * Setup automatic vulnerability scanning based on server-configured interval.
    * Fetches settings from server and sets up recurring scans.
    * Also sets up polling for on-demand scan requests from the server.
    */
   async setupAutoAudit() {
-    const settings = await this.fetchProjectSettings();
+    const settings = await this.fetchProjectSettingsWithRetry();
+    console.log("[MonitorClient] Polling for scan requests enabled (every 5 minutes)");
+    this.settingsPollingTimer = setInterval(() => {
+      this.checkForScanRequest().catch((err) => {
+        this.reportError("SETTINGS_POLL", "Scan request check failed", err);
+      });
+    }, CONFIG_LIMITS.SETTINGS_POLL_INTERVAL_MS);
     if (!settings) {
-      console.warn("[MonitorClient] Could not fetch project settings, auto audit disabled");
+      const message = "Could not fetch project settings after retrying at startup; will keep retrying every 5 minutes";
+      console.warn(`[MonitorClient] ${message}`);
+      this.reportError("AUDIT_SETUP", message);
       return;
     }
     if (settings.scanRequestedAt) {
@@ -479,12 +515,6 @@ var MonitorClient = class {
       await this.runScanAndTrackTime();
       this.setAuditInterval(intervalHours);
     }
-    console.log("[MonitorClient] Polling for scan requests enabled (every 5 minutes)");
-    this.settingsPollingTimer = setInterval(() => {
-      this.checkForScanRequest().catch((err) => {
-        this.reportError("SETTINGS_POLL", "Scan request check failed", err);
-      });
-    }, CONFIG_LIMITS.SETTINGS_POLL_INTERVAL_MS);
   }
   /**
    * (Re)configure the recurring audit timer for the given interval, replacing
@@ -1225,7 +1255,9 @@ var MonitorClient = class {
           maxBuffer: CONFIG_LIMITS.AUDIT_MAX_BUFFER
         });
         if (result2.timedOut) {
-          console.error(`[MonitorClient] yarn audit timed out after ${this.auditTimeoutMs}ms`);
+          const message = `yarn audit timed out after ${this.auditTimeoutMs}ms`;
+          console.error(`[MonitorClient] ${message}`);
+          this.reportError("VULNERABILITY_SCAN", message);
           return null;
         }
         auditOutput = result2.stdout;
@@ -1258,7 +1290,9 @@ var MonitorClient = class {
           maxBuffer: CONFIG_LIMITS.AUDIT_MAX_BUFFER
         });
         if (result2.timedOut) {
-          console.error(`[MonitorClient] npm audit timed out after ${this.auditTimeoutMs}ms`);
+          const message = `npm audit timed out after ${this.auditTimeoutMs}ms`;
+          console.error(`[MonitorClient] ${message}`);
+          this.reportError("VULNERABILITY_SCAN", message);
           return null;
         }
         auditOutput = result2.stdout;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ceon-oy/monitor-sdk",
-  "version": "1.5.3",
+  "version": "1.5.7",
   "description": "Client SDK for Ceon Monitor - Error tracking, health monitoring, security events, and vulnerability scanning",
   "author": "Ceon",
   "license": "MIT",