Intelligent Log Detection¶
Based on intelligent detection algorithms, monitor log data generated by collectors within the workspace. Automatically identify anomalies such as sudden increases or decreases in log volumes and sudden increases in error logs to promptly detect unexpected abnormal states.
Use Cases¶
This is particularly suitable for IT monitoring scenarios involving code exceptions or task scheduling detection. For example, monitoring a sudden increase in error counts in logs.
Detection Configuration¶
-
Define the monitor name.
-
Select the detection dimension: Supports By Source or By Service detection, automatically matching the selected dimension based on user choice;
-
Choose the detection scope: Filter the detection metric data using labels associated with metrics, limiting the scope of detected data. You can add one or more label filters. If no filter is added, all log data will be detected.
View Events¶
The monitor retrieves the last 10 minutes of log metrics. When it detects sudden increases or decreases in log volumes or sudden increases in error logs, corresponding events are generated. These events can be viewed under Events > Intelligent Monitoring.
Event Details Page¶
Clicking Events allows you to view the details page of intelligent monitoring events, including event status, anomaly occurrence time, anomaly name, analysis report, alert notifications, historical records, and related events.
-
Click the Go to Monitor button at the top right to adjust the intelligent monitor configuration;
-
Click the Export button at the top right to choose between exporting a JSON file and exporting a PDF file, thereby obtaining all key data related to the current event.
Analysis Report
-
Anomaly Summary: View the current anomaly log tags, detailed anomaly analysis reports, and distribution of failed request counts.
-
Error Analysis: View clustering information for error logs.
Note: When multiple anomaly intervals exist, the Anomaly Summary > Anomaly Value Distribution Chart and Anomaly Analysis dashboard default to displaying the analysis of the first anomaly interval. You can switch by clicking on the Anomaly Value Distribution Chart, after which the anomaly analysis dashboard will synchronize and update accordingly.