> ## Documentation Index
> Fetch the complete documentation index at: https://docs.flashduty.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Engine disconnection alert

> Configure engine disconnection alert rules to receive timely notifications when the alert engine goes down

If the alert engine (`monitedge`) goes down, alert rules cannot execute, which has a significant impact. The engine disconnection alert feature sends timely notifications when the engine goes down, ensuring the reliability of your monitoring system.

**Menu Entry**: Alert Engine → Engine Disconnection Alert

<Note>
  For an engine cluster composed of multiple instances, as long as at least one instance in the cluster is alive, the disconnection alert will not trigger. It only triggers when all instances in the cluster are disconnected.
</Note>

## Alert rule list

The list displays all configured engine disconnection alert rules, with support for keyword search and customizable display columns.

| Column                         | Description                                                                   |
| ------------------------------ | ----------------------------------------------------------------------------- |
| **Rule Name**                  | Name of the alert rule                                                        |
| **Severity**                   | Critical (red), Warning (orange), Info (yellow)                               |
| **Include Pattern**            | Engine cluster name patterns to monitor; supports wildcard `*`                |
| **Exclude Pattern**            | Engine cluster name patterns to exclude from monitoring                       |
| **Lost Duration (seconds)**    | Triggers an alert after the engine cluster has no heartbeat for this duration |
| **Channel**                    | The channel where alert events are delivered                                  |
| **Trigger Count**              | Maximum number of alert events to generate during disconnection               |
| **Trigger Interval (seconds)** | Time interval between repeated alert event generation                         |
| **Enabled**                    | Enable/disable toggle for the rule                                            |

## Create an alert rule

Click the **New** button to configure the following parameters in the side drawer:

<Steps>
  <Step title="Basic information">
    | Config Item   | Description                             | Default          |
    | ------------- | --------------------------------------- | ---------------- |
    | **Rule Name** | Rule name for identification and search | `monitedge lost` |
    | **Enabled**   | Whether to enable this rule immediately | Enabled          |
  </Step>

  <Step title="Severity">
    Select the severity level for alert events:

    * **Critical**: Urgent, typically used for core engines
    * **Warning**: Warning level, the default
    * **Info**: Informational level
  </Step>

  <Step title="Include and exclude rules">
    | Config Item         | Description                                                                              | Default |
    | ------------------- | ---------------------------------------------------------------------------------------- | ------- |
    | **Include Pattern** | Enter engine cluster name patterns to monitor; supports multiple values; `*` matches all | `*`     |
    | **Exclude Pattern** | Enter engine cluster name patterns to exclude                                            | Empty   |

    <Note>
      Include pattern and exclude pattern cannot both be empty.
    </Note>
  </Step>

  <Step title="Trigger conditions">
    | Config Item                    | Description                                                                                        | Default |
    | ------------------------------ | -------------------------------------------------------------------------------------------------- | ------- |
    | **Lost Duration (seconds)**    | Triggers an alert when all instances in the cluster are disconnected for longer than this duration | 120     |
    | **Trigger Count**              | Maximum number of alert events to repeatedly generate during continuous disconnection              | 3       |
    | **Trigger Interval (seconds)** | Minimum time interval between each repeated alert event                                            | 300     |
  </Step>

  <Step title="Channel">
    Select the channel where alert events should be delivered. Alert events are processed through Flashduty On-call channels for subsequent notification dispatch and handling.
  </Step>
</Steps>

## Edit and delete

* **Edit**: Click the **Edit** button in the list to modify rule configuration. Only the rule creator, primary account, or admin roles can edit.
* **Delete**: Click the **Delete** button in the list. Only the rule creator, primary account, or admin roles can delete.
* **Enable/Disable**: Quickly toggle the rule status using the switch in the list.