The service alarm uses earliest time and deepest node to determine root cause. However, it does not take criticality into account. Here are two alarms, for example, that are combined into a single service alarm with item 1 as the root cause:
1. Major - Blade status unknown
2. Critical - Chassis down
The chassis down actually causes the blade status unknown. This is the more critical alarm that needs to bubble to the top. In addition to earliest/deepest, criticality should be a criteria.
Release : 20.2
Component : CA DOI Foundations
US743141 Criticality is not taken into account for creating the root cause service alarm