Application Real-Time Monitoring Service:Interval detection

Scenarios

• Web & HTML5 application: You can configure static thresholds for web & HTML5 applications to monitor key performance metrics, such as Largest Contentful Paint (LCP), First Input Delay (FID), and Cumulative Layout Shift (CLS). If these metrics encounter anomalies, alerts are triggered to help you quickly locate and fix the anomalies, and improve user experience.

• iOS and Android apps: You can configure static thresholds for iOS and Android apps to monitor key performance metrics, such as resource loading, API calls, crashes, and janks. Automatic reporting of abnormal stacks and crash logs helps you quickly locate and fix the anomalies and improve user experience.

Example:

After a new version of an app was released, the number of application janks significantly increased. In this case, you can configure a static threshold for the number of janks so that you can identify and fix them promptly, thus preventing more users from being affected.

Prerequisites

The application is monitored in RUM. For more information, see Integrate applications.

Configure interval detection

1. Log on to the ARMS console.

2. In the left-side navigation pane, choose Real User Monitoring > Alert rules, and then click Create Alert Rule.

3. On the Create Alert Rule page, enter an Alert Name, and set Alert Detection Type to Interval Detection.

4. In the Alert Contact section, select the application, metric type, and filter conditions.

   In the Alert Rule section, set the Alert Triggering Mode (for example, Single Condition). Then, configure the alert condition, including the detection period (for example, the last 5 minutes), metric (for example, Exception Count), aggregation method (for example, Average), and condition (for example, Outside the dynamic threshold band). Set the alert level (for example, P1: critical) and adjust the tolerance. The system displays a preview of the metric trend and the dynamic threshold band in a line chart. From the Notification Policy drop-down list, select a notification method. After you complete the configuration, click Done.

   Parameter	Description
   Alert Application	Select the application for which you want to create the alert rule. Interval Detection supports only one alert rule per application.
   Metric Type	Select the metric type to monitor. For more information, see RUM alert metrics. After you select a metric type, the system automatically calculates the upper and lower boundaries and renders them in real time. You can preview the metric trend in the Alert Condition section. Note The available condition fields and Filter Conditions depend on the selected Metric Type. The initial rendering may take 2 to 4 seconds. For information about how the upper and lower boundaries are calculated, see How dynamic thresholds work.
   Filter Conditions	Filters the metric to narrow the monitoring scope. You can use the following dimension operators for alert metrics: Traverse: Matches all alerts for the current dimension. = / !=: Specifies one or more values for an exact match or exclusion. Contains / Does not contain: Specifies one or more values for a fuzzy match or exclusion. Note To include dimension information in alert notifications, you must specify the corresponding filter condition. Otherwise, the dimension is aggregated in the metric query results.

5. In the Alert rules section, set the Alert Condition.

   Parameter	Description
   Alert Triggering Mode	Interval Detection supports only single-condition alerts. It does not support combinations of multiple trigger modes.
   Alert Condition	Configure the alert trigger condition based on the selected metric type. Detection Period: Select the detection window, for example, the last 5 minutes. Alert Metric: Select the corresponding metric. Trigger Condition: Specify the trigger condition, such as Greater Than or Equal To, Less Than or Equal To, Month-on-Month Increase/Decrease %, Hour-on-Hour Increase/Decrease %, or Day-on-Day Increase/Decrease %. In the data preview area, the blue line represents the actual data points, and the green area represents the range between the upper and lower boundaries. Note You can evaluate if the current threshold configuration is appropriate and check for past triggers by reviewing alert notifications. For more information, see View alert details.
   Tolerance	Adjusts the width of the band based on the automatically calculated boundaries. A higher tolerance (move the slider to the right) creates a wider band, which makes the model less sensitive to metric fluctuations and triggers fewer alerts. A lower tolerance (move the slider to the left) creates a narrower band, which makes the model more sensitive and triggers more alerts.
   Alert Level	The system provides four built-in alert levels. You can configure different notification policies for each level. For example, you can configure P1 alerts to be sent by phone call and P2 or lower alerts to be sent to a DingTalk group chat. P1: critical P2: error P3: warning P4: info
   Alert Quantity Prediction	Displays the number of times the metric is predicted to exceed the threshold within the selected time period. You can click a specific alert count to view the metric values that would have triggered an alert at historical time points. We recommend using the alert quantity prediction feature when creating or modifying an alert rule. It analyzes historical data to forecast alerts, helping you fine-tune the threshold. For more information, see Alert quantity prediction.

6. Configure the Notification Policy and Advanced Alert Settings.

   Parameter	Description
   Notification Policy	Do not specify a notification policy: When an alert is triggered, no notification is sent. Notifications are sent only when a matching rule in a notification policy is triggered. Specify a notification policy: When an alert is triggered, ARMS sends a notification by using the method defined in the specified notification policy. You can select an existing notification policy or create a new one. For more information, see Notification policies.
   Advanced Alert Settings
   No Data	This setting defines how to handle periods when no metric data is reported. You can configure the system to automatically resolve the alert by treating the value as 0 or 1, or to not trigger an alert at all. For more information, see Glossary for Alert Management.

7. After you complete the settings, click Save.

View alert details

After you receive an alert notification, you can view the alert details in a or on the Alert Management > Alert Event History page of the ARMS console.

How the threshold range is calculated

The dynamic thresholds feature in ARMS primarily uses the Prophet algorithm. After you configure a dynamic thresholds task, ARMS runs in the background once every 24 hours to learn from the metric's historical data from the past 7 days. It extracts features such as trends and seasonality to generate a prediction curve for the next 24 hours. Based on the metric's own fluctuations (error variance), it estimates a range for the metric's data for the next day. When you configure a dynamic thresholds alert, you can preview the upper and lower boundaries calculated by the algorithm.

Unlike static threshold recommendations, you do not need to manually update the alert rule when business changes cause the metric's baseline to shift. This is because ARMS continuously learns the metric's characteristics once a day and predicts boundaries for only the next day. Therefore, you do not need to manually adjust the threshold multiple times.

Alert quantity prediction

The alert quantity prediction feature analyzes historical data to forecast the number of alerts that would have been triggered within a selected period and shows the exact time each alert would have occurred.