Grafana dashboards for Cloudera AI Inference

Cloudera AI Inference service provides dedicated Grafana dashboards that expose model-serving infrastructure metrics. These dashboards offer observability with Cloudera AI Inference service views and expanded metric coverage, enabling easier monitoring and troubleshooting of the Cloudera AI Inference service deployment.

Two types of dashboards are created for the Cloudera AI Inference service:

• Cluster dashboards - They display the metrics for all inference-service related namespaces.
• Container dashboards - They display the metrics corresponding to all the containers belonging to the Cloudera AI Inference service across all namespaces.

The dedicated Cloudera AI Inference service dashboards use the following naming conventions:

• Cluster dashboards: [***ENVIRONMENT NAME***]-[INFERENCE SERVICE NAME***]-cluster
• Container dashboards: [***ENVIRONMENT NAME***]-[INFERENCE SERVICE NAME***]-containers

The Cloudera AI Inference service dashboards provide clearer, more accurate monitoring of inference workloads and simplify troubleshooting by presenting only the metrics that are relevant to Cloudera AI Inference service.