AIAPM Alerts

The AIAPM Alerts section under IT Management > Alarms/Triggers provides a detailed view of anomalies detected by the AI-based Application Performance Monitoring system. Displays alerts related to latency, error rates, and performance anomalies in critical services.

Alert Lifecycle

Alert Generation

1. Monitoring Phase: System continuously monitors your application metrics
2. Threshold Detection: When a metric exceeds your defined threshold, monitoring phase begins
3. Persistence Check: If the issue persists beyond your defined interval, an alert is generated
4. Initial Notification: WARNING level alert sent to configured recipients

Alert Escalation

1. Continued Monitoring: System continues to track the issue
2. Frequency Check: If the issue persists beyond the frequency period, the alert escalates
3. Critical Notification: CRITICAL level alert sent to configured recipients
4. Assignment: Alerts can be manually assigned or auto-assigned based on your configuration

Alert Resolution

1. Recovery Detection: When metrics return below threshold, system starts recovery monitoring
2. Automatic Resolution: If metrics remain normal, the alert is automatically resolved
3. Resolution Notification: A resolution notification is sent to original recipients
4. Documentation: All alert details are preserved for historical reporting and analysis

Best Practices

• Set thresholds based on your application's normal behavior
• Configure shorter intervals for critical services
• Establish a clear workflow for alert handling within your team
• Use the comment feature to document troubleshooting steps
• Review historical alerts to identify patterns and potential system improvements

The AIAPM Alerts table provides a comprehensive view of all anomalies detected by the AI-powered Application Performance Monitoring system. Each row in the table represents a unique alert instance, starting with a serial No for reference. The Create Time column logs the exact timestamp when the alert was generated, helping track when issues occurred. The Severity column indicates the criticality of the issue, typically marked as Warning. All alerts under this section are categorized under the Trigger Type as AIAPM Anomalies. The Description provides specific details about the anomaly, such as the impacted application or service. The Type column shows whether the alert has been addressed, commonly labeled as Not Resolved, while the Status column reflects the nature of the problem, such as Problem, to indicate ongoing issues. This table enables users to quickly assess, prioritize, and respond to performance anomalies within their systems. 

Managing Alerts:

• Users can select one or more alerts using the checkbox in the first column.

• After selecting, click on the Actions dropdown button to choose one of the following:

  • Completed – Marks the alert as resolved and completed.

  • Acknowledged – Indicates the alert has been reviewed.

  • Move to History – Archives the alert for historical reference and removes it from the active list.

Exporting Alert Data:

• Click the Export button on the top right to download the list of anomalies.

• The exported file helps with offline analysis, reporting, or sharing with external teams.

Threshold Notification Management

When a user clicks on any alert entry from the AIAPM Alerts table, a pop-up window titled Threshold Notification Management is displayed. This window allows users to document the alert response, track investigation efforts, and mark the alert status accordingly.

Popup Fields & Sections:

• Created Date: Displays the timestamp of when the alert was generated.

• Description: Shows the anomaly message related to the specific application or service.

 Root Cause Analysis (RCA):

• This section is expanded to document the technical investigation into the root cause of the alert (e.g., code issue, service delay, infrastructure bottleneck).

Span Details:

• The Span Details section in the Threshold Notification Management popup provides a detailed view of the specific operation that triggered the AIAPM alert. It includes technical attributes such as the operation name, type, and subtype, the specific action performed, and the duration of execution. Additional fields like id, db, and transaction offer insights into the database involved, trace identifiers, and the broader transaction context. 

• Taken By *: Dropdown to select the user who has taken ownership of the alert resolution.

• Time Required *: Field to specify the estimated time (in hours) needed to resolve the issue.

• Comments *: A mandatory field for entering relevant notes or observations (e.g., reason for anomaly, mitigation steps).

All fields marked with an asterisk (*) are mandatory to proceed.

Two radio button options are provided to set the current state of the alert:

• Acknowledge: Marks the alert as seen and under review.

• Closed: Marks the alert as fully resolved and completed.

• Submit: Saves all entered details and updates the alert record.