Workbenches
Cloudera Observability provides a centralized dashboard for managing and monitoring AI resources across your environment. By navigating to the Real Time tab, you can gain immediate visibility into the health and utilization of your infrastructure. The interface allows you to track specific workbenches and AI workloads, ensuring that you have the necessary oversight to maintain performance and scale your AI operations effectively.
This platform enables you to monitor active users and the busyness of individual workbenches to optimize resource allocation. Cloudera AI helps you identify which workbenches require additional nodes or where workloads might be underutilizing assigned resources.
You can drill down into specific workbench details to understand the relationship between nodes, projects, and active AI workloads.
The Workbenches table provides detailed status for each active environment.
| Column | Description |
|---|---|
| Workbench | Displays the unique name of the workbench instance. |
| Version | Indicates the specific software version currently running on the workbench. |
| Nodes | Lists the number of nodes allocated to the workbench. |
| Projects | Shows the total number of projects associated with the workbench. |
| AI Workloads | Displays the count of active Cloudera AI workloads currently running. |
| Active Users | Indicates the number of users currently engaged with the workbench. |
| Busyness | Provides a percentage representing the current resource utilization of the workbench. |
Resource utilization across workbenches
- Memory - Displays the memory usage in GiB.
- CPU - Displays the CPU usage in %.
- GPU - Displays the GPU usage in %.
- Busyness Activity - A line chart displays the average consumption of resources for the selected workbench, illustrated as a percentage. This metric indicates whether you over-allocate or under-allocate resources to the selected workbench.
- Network input and output - Displays the duration and volume of data sent and received over the network during processing.
- Data Read and Write - Indicates the time and volume of data read from and written to the engine's storage.
