New type of policies that replace the Temporal Triggers in Netcool/OMNIBus

Background

During the migration of Aggregation ObjectServer triggers from Netcool/OMNIbus to IBM AIOps, a critical gap has been identified. The existing temporal triggers in OMNIbus—which execute periodically after a defined time interval to evaluate existing alerts against specific conditions—currently have no direct equivalent in the AIOps Policy framework.

Problem Statement

In OMNIbus, temporal triggers play an essential role in ongoing alert management by:

• Re-evaluating active alerts over time (e.g., escalating alerts not cleared within a set duration).
• Updating alert states or severities based on elapsed time or other evolving conditions.
• Performing housekeeping tasks (e.g., clearing stale alerts or re-synchronizing statuses).

When migrating to AIOps, this functionality is not natively supported, leading to a significant gap in automation and event lifecycle management. Current AIOps policies are reactive—triggered by incoming events—rather than proactive or time-based.

Proposed Enhancement

Introduce a new Policy Type: “Scheduled or Temporal Policy” within IBM AIOps that replicates and extends the functionality of temporal triggers.

Key Capabilities:

1. Time-based Execution:
   1. Allow policies to run at fixed intervals (e.g., every 5 minutes, hourly, daily).
   2. Support cron-like scheduling syntax for flexible timing.
2. Alert Querying Scope:
   1. Policies can query existing alerts from the AIOps event database or topology service.
   2. Define filters (e.g., severity, status, duration) to target relevant alerts.
3. Conditional Actions:
   1. Perform actions based on time and state conditions (e.g., “if an alert remains Critical for more than 30 minutes, escalate or notify”).
4. Integration with Existing Policies:
   1. Reuse existing AIOps policy actions (runbook, notification, enrichment, or event update).
   2. Maintain consistent execution and logging framework.