NVIDIA NICo Hardware Health

NICo Hardware Health collects hardware health metrics and BMC logs from configured Redfish targets. Fleet Intelligence Out-of-Band Collector deployments use this service alongside the inventory and attestation collector to monitor hardware through BMC management interfaces.

Features and use cases

• Collect hardware health metrics and BMC logs from supported Redfish endpoints.
• Send telemetry through the deployment's OpenTelemetry gateway.
• Use target and credential configuration generated by the OOB collector deployment.
• Monitor hardware independently of the host operating system.

Requirements and deployment

Use a compatible Linux container runtime. Provide the NICo configuration, BMC credentials, network access to the target BMCs, and connectivity to the telemetry gateway. The OOB Helm chart generates the NICo configuration and mounts credentials through Kubernetes Secrets.

Source and licensing

Published by NVIDIA. This image builds the hardware health service from NVIDIA infra-controller, licensed under Apache License 2.0. License and third-party notices are included under /licenses in the image.

Documentation

See the OOB installation guide, configuration reference, and NICo image build definition.