Reports
You can either use predefined reports as-is or modify them to meet your needs. You can also create various types of reports related to integration entities. All reports can be accessed from the Reports tab.
After a report is saved, the report can be persistently viewed from the Dashboard tab in IO.
For more information, see “Reports” in the Virtana Infrastructure Observability User Guide.
Predefined Reports List
The following predefined reports are provided for this Integration and can be viewed in the IO Dashboard.
Name | Description |
AI Factory Overview | Fleet-wide GPU utilization, energy, errors, and throttling across an AI factory |
AI Job Report | AIFO Job reports |
AI Model - NIM Token Overview | NIM inference request, token, and latency metrics for an AI model |
AIDC Program: $entity_id | GPU utilization and application details for a program |
Application_GPU_Utilization | GPU, memory, and power usage, and NVLink connectivity, for an application |
Cluster GPU - Network | Network and storage throughput across a GPU cluster |
Cluster GPU - Overview | Cluster-wide GPU utilization, errors, energy, and storage summary |
GPU Cluster - Dashboard | Cluster-wide dashboard for GPU utilization, throttling, errors, and storage |
GPU Efficiency - GPU | GPU efficiency, wasted cycles, and energy consumption for a GPU |
GPU Efficiency - Host | GPU efficiency, wasted cycles, and energy consumption for a host |
GPU Energy Overview | GPU power, temperature, and throttling trends |
GPU Entity - Program Report | Programs and applications running on a GPU |
GPU Errors | ECC and PCI error counts and trends for a GPU |
Host Energy Overview | GPU power, temperature, clock rate, and throttling for a host |
Host GPU Errors | ECC and PCI error counts by GPU for a host |
Host GPU Overview New | GPU and vGPU utilization, throttling, and memory usage for a host |
Host GPU Programs | Programs that have run on a host's GPUs |
NVIDIA Host - Network | Network switch utilization and connectivity for a host |
NVIDIA Host - Storage | PowerScale node latency, CPU usage, and throughput for a host |
NVIDIA Host GPU Configuration | GPU device configuration and connectivity details for a host |
NVIDIA Host NVLink Configuration | NVLink topology and byte-rate details for a host's GPUs |
NVIDIA Overview | Overview of an NVIDIA host, its GPUs, and programs |
Template - Hitachi UCP Overview | Showcases the capabilities of Hitachi UCP |