Dashboard metrics

Reference for the HCL Universal Orchestrator dashboard metrics, including metric names, filtering labels, aggregation types, and technical definitions.

Metrics provide insights into workload health and resource utilization. By default, these metrics are collected from system services every 20 seconds, but you can customize this value by editing the uno.metrics.default.sent.frequency parameter in the values.yaml file.

Each metric is identified by a name and might have labels that enable you to filter data, for example by folder and workstation.

Access to these metrics is governed by your security profile. To view data for an item, you must have the LIST permission for that specific item.

The below table displays the core metric definitions:
Table 1. Dashboard metrics
Metric name Description Filtering labels Default aggregation
Workstation limit The numerical task capacity or threshold defined for a specific workstation.
Note: For the Workstation limit metric, workstations are not displayed unless at least one task has been submitted to them.
 folder, name Average
Workstation linked Indicates if a workstation is currently connected and active within the plan:
  • 1: the workstation is linked.
  • 0: the workstation is not linked.
 folder, name Sum
Tasks by status The total number of tasks in the plan, grouped by their folder or current status, for example successful, error, waiting. folder, taskstatus Sum
Daily task consumption for license monitoring The count of tasks or AI agent calls processed by day, used primarily for monitoring license consumption. type Sum
Monthly task consumption for license monitoring The count of tasks or AI agent calls processed by month, used primarily for monitoring license consumption. type Sum
Total task consumption for license monitoring The total count of tasks or AI agent calls, used for license tracking. type Sum
Task run delay by workstation The amount of latency in milliseconds occurring during the run of a task, by workstation. folder, name Average
Submitted task rate The number of task submissions occurring over a specific time interval. folder Sum
Tasks with deadline exceeded The total count of tasks that exceeded their scheduled deadline. folder Sum
Tasks with maximum duration exceeded The total count of tasks that exceeded their required maximum duration. folder Sum
Tasks with minimum duration not reached The total count of tasks that completed before reaching their required minimum duration. folder Sum
v2.1.5+GenAI unit consumption within the last 24 hours The count of GenAI units processed within the last 24 hours. model, type Sum
v2.1.5+GenAI unit consumption within the last 30 days The count of GenAI units processed within the last 30 days. model, type Sum
v2.1.5+AI agent calls within the last 24 hours The count of AI agent calls processed within the last 24 hours. model Sum
v2.1.5+AI agent calls within the last 30 days The count of AI agent calls processed within the last 30 days. model Sum
v2.1.5+AI agent response delay within the last 24 hours The average latency in milliseconds occurring during an AI agent call within the last 24 hours. model Average
v2.1.5+AI agent response delay within the last 30 days The average latency in milliseconds occurring during an AI agent call within the last 30 days. model Average
v2.1.5+Estimated GenAI unit cost for the last 24 hours The cost estimation for GenAI units processed within the last 24 hours. model Sum
v2.1.5+Estimated GenAI unit cost for the last 30 days The cost estimation for GenAI units processed within the last 30 days. model Sum
Metric behavior and logic
Aggregation and visualization
The way data is summarized depends on the metric type:

  • Sum: Used for counts like Submitted task rate or Tasks by status to show totals.

  • Average: Used for performance indicators like Task run delay by workstation or Workstation limit to show steady-state behavior.

To modify the aggregation type of a metric, select an option from the Aggregation menu within the Configuration tab of the widget panel.
Note: By default, metrics are aggregated into a single total. To see individual breakdowns, use the Group By setting in the widget configuration tab.
Data freshness
Metrics are polled every 20 seconds.
  • When viewing a Gauge widget, the displayed data reflects the value of the most recent 20-second interval.
  • When viewing a Time series widget, the line reflects the trend of the intervals over your selected time range.
  • v2.1.5+When viewing Pie chart or Bar chart widgets, the displayed data reflects the value of the most recent 20-second interval.
When viewing a Gauge widget, the needle reflects the value of the most recent 20-second interval. When viewing a Time series widget, the line reflects the trend of the intervals over your selected time range.
Data retention
By default, the dashboard displays historical data for up to 15 days. Information older than this period is automatically removed from the timeseries database to ensure optimal system performance.
You can modify data retention values under Settings > System > Timings and expiration. For more information, see Timings and expiration.