Skip to main content
Version: v3.8.0

Step 9 - Alert dashboard Service

Goal​

Create the alert-dashboard Service so operators reach the dashboard through Router. Complete Phase 1 verification for the Smart Factory story.

Scenario context​

alert-dashboard listens on external port 8082 on ops-b. The Service publishes a mesh bridgePort for browser access at:

http://router.default.svc.bridge.local:<dashboard-bridge-port>/

This is the primary UI for Phase 1: fleet sparklines, live alerts, event journal, and Diagnose actions.

YAML walkthrough​

File: deploy/steps/07-service-alert-dashboard.yaml

alert-dashboard Service
---
apiVersion: datasance.com/v3
kind: Service
metadata:
name: alert-dashboard
tags:
  - pot-edge-patterns
spec:
type: microservice
resource: operations-center/alert-dashboard
targetPort: 8082
defaultBridge: default-router

After deploy, run potctl describe service alert-dashboard and note bridgePort for your browser URL.

Optional step 10: direct agent IP access​

For lab demos without Router DNS, set ports.external on the alert-dashboard microservice in step 8 YAML so the dashboard also listens on a host port on ops-b. This is optional. Router access is the recommended path for production-style demos.

Why this design​

The dashboard is HTTP like diagnostic-service, but it runs on ops-b while consuming NATS data imported from site-a. The Service bridge gives presenters one stable URL on the Router mesh.

Deploy​

potctl deploy -f deploy/steps/07-service-alert-dashboard.yaml

Verify​

Service and browser​

potctl describe service alert-dashboard

Open http://router.default.svc.bridge.local:<dashboard-bridge-port>/ in a browser.

Phase 1 checklist​

  1. Fleet panel: sparklines for M001–M004 updating from telemetry.machine.{machineId}
  2. Live alerts: set REPLAY_SEGMENT=fault-high-vibration-m002 and REPLAY_SPEED=5 on sensor-simulator (redeploy step 3) or wait for dataset fault rows; alert shows faultType
  3. Event journal: same alert persists after page refresh (JetStream path)
  4. Diagnose: click an alert row; runbook JSON returns via DIAGNOSTIC_URL

Log checks​

potctl describe microservice sensor-simulator
potctl describe microservice smart-anomaly-detector

Sensor logs should show four machines publishing. Detector logs should show subscription to telemetry.machine.>.

Common mistakes​

  • Empty live alerts. Re-check step 4 account key and step 5 import rule.
  • Empty fleet charts. Confirm fleet-telemetry export (step 2) and import (step 5), plus TELEMETRY_SUBJECT on dashboard.
  • Journal empty but live works. Verify JetStream export/import and detector stream create logs.
  • Diagnose 502. Wrong DIAGNOSTIC_URL port; redeploy step 8 with bridgePort from step 7.
  • Using v1.0.0 images. Redeploy all Applications with :1.1.0 and matching subject names.

Next step​

Phase 1 is complete. Continue with Step 11: Vision NATS user rules for fab wafer inspection, or return to the Acme Smart Plant overview verification checklist.

Group 3See anything wrong with the document? Help us improve it!

Edit this page on GitHub!