Skip to navigation

NIM Service Chart

View as MarkdownOpen in Claude

The nvidia-bnr-h4m-service Helm chart deploys the BNR NIM as a standalone inference service on Kubernetes or Red Hat OpenShift.

  • Input: Audio stream via SMPTE ST 2110-30.
  • Output: Enhanced audio via SMPTE ST 2110-30.

This chart can be deployed standalone for integration into a custom media pipeline, or as a subchart of the End-to-End Demo Chart.


Installation

Prerequisites

Complete all prerequisite steps (Rivermax license, image pull and model pull secrets, high-speed network attachment) before running helm install. For details, refer to Getting Started.

Pull the Chart

If not already done, add the Helm repository and pull the chart.

helm pull nim-repo/nvidia-bnr-h4m-service --version 2.0.0

For the full repository setup, refer to Pull Helm Charts.

The downloaded .tgz is referenced directly in the following installation commands.

Helm Installation

Default deployment (NMOS mode):

helm upgrade --install nvidia-bnr-h4m-service \
nvidia-bnr-h4m-service-2.0.0.tgz

With node hostname, network, scheduler, and secret overrides:

helm upgrade --install nvidia-bnr-h4m-service \
nvidia-bnr-h4m-service-2.0.0.tgz \
--set nodeSelector.hostname=<gpu-node-name> \
--set image.secret=<image-pull-secret> \
--set schedulerName=<scheduler-name> \
--set network.name=<multus-net-attach-def> \
--set ngc.secretName=<model-pull-secret>
--set FlagPurposeDefault
nodeSelector.hostnameTarget GPU node.example-gpu-node
image.secretImage pull secret (docker-registry type) for nvcr.io.ngc-image-secret
schedulerNamePod spec.schedulerName.topo-aware-scheduler
network.nameMultus NetworkAttachmentDefinition for SMPTE ST 2110.media-a-tx-net
ngc.secretNameModel pull secret (generic type) for NGC model download.ngc-api-key

The chart also creates an endpoint Service named <appName>-endpoint. It exposes the NIM HTTP API on port 8000. When nmos.enabled=true, it also exposes the NMOS Node API on port 9010.

After installation, check the status:

kubectl get pods -o wide
kubectl rollout status deployment/<appName> --timeout=300s

On Red Hat OpenShift, replace kubectl with oc.


Deployment Modes

AspectSMPTE ST 2110 (static)NMOS
Values filevalues-st2110.yamlvalues-nmos.yaml
Stream discoveryFixed multicast IP addresses and ports in values.Dynamic discovery via NMOS Connection Manager.
Use caseKnown network, fixed topology.Dynamic registration and connection management.

SMPTE ST 2110 Static Mode

helm upgrade --install nvidia-bnr-h4m-service \
nvidia-bnr-h4m-service-2.0.0.tgz \
-f values-st2110.yaml

NMOS Mode

helm upgrade --install nvidia-bnr-h4m-service \
nvidia-bnr-h4m-service-2.0.0.tgz \
-f values-nmos.yaml

Starter Configuration Files

Copy the appropriate configuration below into values-st2110.yaml or values-nmos.yaml, adjust the fields for your environment, and pass it to helm upgrade --install with -f.

ST 2110 Configuration - Static (values-st2110.yaml)

appName: nvidia-bnr-nim-st2110
replicas: 1
nodeSelector:
hostname: example-gpu-node
schedulerName: topo-aware-scheduler
image:
repository: nvcr.io/nim/nvidia/bnr-h4m
tag: "2.0.0"
pullPolicy: IfNotPresent
secret: ngc-image-secret
nmos:
enabled: false
httpPort: 9010
seed: nvidia-bnr-nim-st2110
hostname: bnr-nim.local
description: BNR-NIM
label: BNR-NIM
service:
enabled: true
type: NodePort
nimHttpPort: 8000
nimHttpNodePort: 32516
model:
type: v2-48k
audio:
pcmFormat: S24BE
numChannels: 1
intensityRatio: 1.0
input:
audio:
sessionName: audio_in
localInterfaceName: net1
hostIp: "234.5.8.9"
hostPort: 5003
hostNumSubnetBits: 24
output:
audio:
sessionName: audio_out
localInterfaceName: net1
hostIp: "234.5.8.9"
hostPort: 5004
hostNumSubnetBits: 24
network:
name: media-a-tx-net
bnr:
nmosEnabled: false
nimEnabled: true

NMOS Configuration (values-nmos.yaml)

appName: nvidia-bnr-nim-nmos
replicas: 1
nodeSelector:
hostname: example-gpu-node
schedulerName: topo-aware-scheduler
image:
repository: nvcr.io/nim/nvidia/bnr-h4m
tag: "2.0.0"
pullPolicy: IfNotPresent
secret: ngc-image-secret
nmos:
enabled: true
httpPort: 9010
seed: nvidia-bnr-nim-nmos
hostname: bnr-nim.local
description: BNR-NIM
label: BNR-NIM
service:
enabled: true
type: NodePort
port: 9010
nodePort: 32515
nimHttpPort: 8000
nimHttpNodePort: 32516
model:
type: v2-48k
audio:
pcmFormat: S24BE
numChannels: 1
intensityRatio: 1.0
input:
audio:
sessionName: audio_in
output:
audio:
sessionName: audio_out
network:
name: media-a-tx-net
bnr:
nmosEnabled: true
nimEnabled: true

The audio sample rate is derived from model.type: v2-48k uses 48000 Hz, and v2-16k uses 16000 Hz. Do not set audio.samplingRate unless you are diagnosing a chart or runtime issue.


Verify

helm status nvidia-bnr-h4m-service
kubectl get pods -o wide

On Red Hat OpenShift, replace kubectl with oc.


Uninstall

helm uninstall nvidia-bnr-h4m-service

The *-model-cache and *-nim-logs PVCs carry helm.sh/resource-policy: keep and are not removed by helm uninstall.