Setting metric threshold values
Acceptable threshold values depend on your operating environment. When a threshold is crossed, the state of the metric score changes.
For example, you determine that the maximum acceptable queue length is 50 items. You select Low values are good. You set the upper value to 50 and the lower value to 40. If the queue remains below 40 items in length, the metric score is green (good). If the queue length goes above 40 items, the metric score is yellow (average). If the queue length goes above 50 items, the metric score is red (poor).
Or for percentage of successful requests, you select High values are good. You set the upper value to 98 and the lower value to 95. If the percentage of successful requests goes below 95 percent, the metric score is red (poor). If the percentage of successful requests is between 95 and 98 percent, the metric score is yellow (average). If the percentage of successful requests remains above 98, the metric score is green (good).
Changes to thresholds are effective immediately.
There are no threshold defaults. You must set thresholds for metric scores to display.
If you want to be notified when thresholds are exceeded, you can create an agent Agents. Sample agents that monitor the audit database for threshold violations and perform common actions when violations are detected are included in the audit samples package.
Before you begin
Log entries Setting up Logging occur in the following circumstances:
- when metric thresholds are violated
- when enumerated metrics, such as operational status, change
Logs are not generated when metric values change but remain in the same range.