Alert events

Updated at:

The Alert Events page displays alert events from modules such as offline computing, real-time computing, real-time workflow, baseline monitoring, Data Quality, and DataService Studio. This topic describes how to view and handle alert events.

View alert events

  1. On the Dataphin page, click the alert ddd icon in the upper-right corner to open the Alert Center page.

    image

  2. On the Alert Event page, click the Offline Computing, Real-time Computing, Real-time Workflow, Baseline Monitoring, Data Quality, or Data Service tab to view the alert events for the corresponding module. The layout is the same for all module tabs. The following figure uses the Offline Computing tab as an example.

    image

    Area

    Description

    ① Filter and search area

    Search for alert events by keywords such as the alert object name. You can also filter alert events by Alert Recipient, Alert Date, Alert Status, File Format, Alert Reason, Business Unit, or Project.

    • Alert Recipient: Filter by yourself (superadmin), all users, the owner of the alert object, an on-call schedule, or custom users.

    • Alert Date: Select an alert date within the last month.

    ② Alert event list

    The list shows the object name, latest alert time, alert recipient, alert reason, project, and current status for each alert event. You can also Mute or Handle Alert for an alert object.

    • Alert Reason: The cause of the alert event. An alert event is triggered by a single alert rule. Alert reasons vary by feature.

      • Offline computing includes Error, Completed, Execution Timeout, and Not Completed.

      • Data Quality includes Abnormal Table Rule, Abnormal Field Rule, and Abnormal Custom Rule.

      • DataService Studio includes Abnormal Average Response Time, Abnormal Number of Calls, Abnormal Error Rate, and Abnormal Offline Percentage.

      • Real-time computing includes High Business Latency, TPS Out of Range, Failure Rate Exceeds Configured Ratio, Data Backlog Exceeds Configured Threshold, and Backpressure Duration Exceeds Configured Threshold.

        The supported alert types vary by compute engine.

        • If a project in a tenant is attached only to a Blink compute engine, Backpressure Duration Exceeds Configured Threshold is not supported.

        • If a project in a tenant is attached only to an open source Flink compute engine, High Business Latency and Data Backlog Exceeds Configured Threshold are not supported.

        • If the real-time computing engine attached to a project in a tenant includes both Blink and open-source Flink compute engines, this is a supported configuration.

      • Real-time workflow includes High Business Latency, Failure Frequency Exceeds Configured Threshold, TPS Out of Range, and Execution Failed.

    • Current Status: The current status of the alert object. Statuses include Alert Resolved, Alerting, Muted, and Alerting (Cool-down Period).

      • Alert Resolved: The alert event is resolved.

      • Alerting: The alert mode is set to frequent alerting, and the alert has started but not yet ended.

      • Muted: You can set alert events with the Alerting or Alert Resolved status to Muted.

      • Alerting (Cool-down Period): This status is based on custom alert logic from different systems. For example, DataService Studio collects metrics every second, but the alert interval is at least 5 minutes.

    • View alert event: Click the name of an alert object to view its details. For more information, see View alert event details.

    • Mute: You can set unimportant alert objects to the Muted status. For more information, see Set the status of an alert event to Muted.

    • Handle alert: Select an alert object to handle. Click the Handle 去处理 icon or click an Object Name. On the Alert Event Details page, click Handle to go to the node page and resolve the alert based on its reason.

View alert event details

On the All Events page, click an Object Name to open the Alert Event Details page. On this page, you can view details such as the object type, source system, alert reason, alert frequency, first alert time, last alert time, number of alerts, alert recipients, and alert method. You can also Mute or Handle Alert for the object, or click related links for more information.

Set the status of an alert event to Muted

You can set unimportant alert events to the Muted status. You can only mute an alert when its status is Alerting or Alert Resolved.

  1. Find the alert event that you want to mute. In the Actions column, click the Mute ddd icon. Alternatively, you can click an Object Name and then click Mute on the Alert Event Details page.

  2. In the Mute Alert dialog box, select a Mute Duration. You can select All Day or enter a custom duration.

  3. Click OK.

    The status of the alert event changes from Alerting or Alert Resolved to Muted.