Use the Alert Manager

This article refers to Platform v3.1.0. The current Platform version is v3.2.0.

Overview

The Alert Manager is a Marketplace application that watches a Barbara edge node and triggers alerts when something goes wrong. It covers three channels of monitoring:

• Service monitoring — periodic health checks of selected workloads on the node, with retries before triggering an alert.
• Telemetry monitoring — every 60 seconds it compares the node's resource usage (notably disk) against user-defined warning and critical thresholds.
• External variables — a REST endpoint that lets your own apps push values to be evaluated against alarm rules.

Alerts are dispatched as email to a configurable recipient list, optionally over MQTT to a broker on the same node, with an optional daily summary email.

Add the Alert Manager to your Library

Alert Manager

Watches services, telemetry, and external variables on an edge node and fires email or MQTT alerts when configured thresholds are breached.

Alert Manager in the Marketplace

For the full Marketplace flow, see Add a Marketplace app to your Library.

Deploy and configure disk monitoring

On the Node Details page of the target node, click Add card → Application and pick Alert Manager to launch the install wizard.

App Secrets

The App Secrets carry the MQTT Broker credentials used when MQTT notifications are enabled. If you only want email alerts, leave the defaults — they are harmless when MQTT is off.

App Secrets in the install wizard

App Config — disk usage thresholds

Paste the following JSON in the App Config step to enable email alerts when disk usage crosses 75 % (warning) and 90 % (critical).

{
  "alarmmanager": {
    "debugLevel": "info",
    "notification": {
      "email_enabled": true,
      "email_recipients": ["user@company.tech"],
      "mqtt_enabled": false,
      "summary_enabled": false,
      "summary_time_utc": "09:00"
    },
    "telemetry": {
      "disk_alarm_warning_threshold": 75,
      "disk_alarm_critical_threshold": 90
    },
    "services": []
  }
}

App Config in the install wizard

Key fields:

Field	Description
notification.email_enabled	Toggle email alerts on/off.
notification.email_recipients	List of email addresses to send alerts to.
notification.mqtt_enabled	Toggle MQTT alerts on/off (requires an MQTT broker on the node).
telemetry.disk_alarm_warning_threshold	Disk usage % that triggers a warning alert.
telemetry.disk_alarm_critical_threshold	Disk usage % that triggers a critical alert.

Click Send App. Alerts now fire whenever disk usage crosses the thresholds.

Monitor a deployed service

Add an entry to the services array of the App Config to watch a workload on the node. The example below watches the Marketplace MQTT Broker (service name mqttbbr) and triggers an alert if its container goes down for more than 30 seconds and one retry:

{
  "alarmmanager": {
    "debugLevel": "info",
    "notification": {
      "email_enabled": true,
      "email_recipients": ["user@company.tech"],
      "mqtt_enabled": true,
      "summary_enabled": true,
      "summary_time_utc": "16:50"
    },
    "telemetry": {
      "disk_alarm_warning_threshold": 40,
      "disk_alarm_critical_threshold": 90
    },
    "services": [
      {
        "enabled": true,
        "serviceName": "mqttbbr",
        "interval": 30,
        "retries": 1
      }
    ]
  }
}

Field	Description
enabled	Toggle this service entry on/off.
serviceName	The container's service name as declared in its compose file.
interval	Seconds between health checks.
retries	Failed checks before the alert fires.

Find the service name

Open the Docker Compose Info segment of any workload card on the Node Details page. The serviceName is the key under services: of the compose file.

Finding the mqttbbr service name in Docker Compose Info

Monitor external variables over REST

The Alert Manager exposes a REST endpoint at http://alarmmanager:5100/event on the node's barbaraServices network. Any workload on the same node — your own or a Marketplace app — can POST evaluation requests to it.

The payload is a list of event objects:

{
  "enabled": false,
  "name": "Low pressure",
  "deviceDisplayName": "machine_01",
  "displayName": "variable_01",
  "value": 20,
  "operation": "<",
  "threshold": 19,
  "retries": 3
}

Example Python client:

import json
import requests

url = "http://alarmmanager:5100/event"

payload = json.dumps([
    {
        "enabled": True,
        "name": "Low pressure",
        "deviceDisplayName": "machine_00",
        "displayName": "variable_01",
        "value": 18,
        "operation": "<",
        "threshold": 19,
        "retries": 3,
    },
    {
        "enabled": False,
        "name": "Low pressure",
        "deviceDisplayName": "machine_01",
        "displayName": "variable_01",
        "value": 20,
        "operation": "<",
        "threshold": 19,
        "retries": 3,
    },
])

response = requests.post(url, headers={"Content-Type": "application/json"}, data=payload)
print(response.text)

See the app's Technical Notes on Marketplace for the full payload reference.

Use with Marketplace industrial connectors

Every industrial connector in the Barbara Marketplace ships pre-integrated with the Alert Manager — you only need to add the alarm configuration to the connector's App Config to start receiving alerts on its variables.

See also...

Browse the connector catalogue and read the per-connector Technical Notes at Industrial Connectors in Barbara Marketplace.

Daily summary email

The Alert Manager can send a once-a-day digest email with every event that occurred in the previous 24 hours. Enable it under notification:

Field	Description
notification.summary_enabled	Enable the daily summary email.
notification.summary_time_utc	Time of day (UTC, HH:MM) when the summary is sent.

Example:

{
  "alarmmanager": {
    "debugLevel": "info",
    "notification": {
      "email_enabled": true,
      "email_recipients": ["user@company.tech"],
      "mqtt_enabled": false,
      "summary_enabled": true,
      "summary_time_utc": "09:00"
    },
    "telemetry": {
      "disk_alarm_warning_threshold": 75,
      "disk_alarm_critical_threshold": 90
    },
    "services": []
  }
}

Summary

The Alert Manager is the catch-all alerting layer for a Barbara node: telemetry thresholds out of the box, service health checks with retries, and a REST endpoint that lets any app on the barbaraServices network push variables for evaluation. Email and MQTT cover the notification side, with an optional daily summary digest.