Customizing Resource Manager health and performance monitoring
- Add and model one or more Resource Manager instances as devices. For each device, specify the device class /ZenossRM, and set configuration properties that have prefix zRMM.
- To determine base-level performance for your Resource Manager systems, study component graphs after a week, a month, and so on. If a metric for a component in one of the pipelines captures data that indicates a potential problem, take the following actions:
To the template for the affected component, add thresholds for the data point.
- Create a threshold to send a warning level event that you can investigate.
- Create another threshold to send an error level event to indicate a more serious problem.
- Create events with event class /ZenossRM for the threshold violations, which will cause a visible change to the related pipeline node.
- As you collect more data, make adjustments to thresholds and events as needed.