NVIDIA Docs Hub NVIDIA Networking Networking Solutions RDG for DPF Zero Trust (DPF-ZT) with OVN VPC DPU service Home DPU Provisioning

DPU Provisioning

Connect to the first DPU BMC over SSH to change the BMC root's password:

Jump Node Console

Copy
Copied!

            
            $ ssh root@10.0.110.201
root@10.0.110.201's password: <BMC Root Password. Default root/0penBmc. need to change first time>

Verify that the rshim service is runnging or run if not.

Jump Node Console

Copy
Copied!

            
            root@dpu-bmc:~# systemctl status rshim
* rshim.service - rshim driver for BlueField SoC
     Loaded: loaded (/usr/lib/systemd/system/rshim.service; enabled; preset: disabled)
     Active: active (running) since Mon 2025-07-14 13:21:34 UTC; 18h ago
       Docs: man:rshim(8)
    Process: 866 ExecStart=/usr/sbin/rshim $OPTIONS (code=exited, status=0/SUCCESS)
   Main PID: 874 (rshim)
        CPU: 2h 43min 44.730s
     CGroup: /system.slice/rshim.service
             `-874 /usr/sbin/rshim
 
Jul 14 13:21:34 dpu-bmc (rshim)[866]: rshim.service: Referenced but unset environment variable evaluates to an empty string: OPTIONS
Jul 14 13:21:34 dpu-bmc rshim[874]: Created PID file: /var/run/rshim.pid
Jul 14 13:21:34 dpu-bmc rshim[874]: USB device detected
Jul 14 13:21:38 dpu-bmc rshim[874]: Probing usb-2.1
Jul 14 13:21:38 dpu-bmc rshim[874]: create rshim usb-2.1
Jul 14 13:21:39 dpu-bmc rshim[874]: rshim0 attached
 
root@dpu-bmc:~# ls /dev/rshim0
boot     console  misc     rshim
 
# To start the rshim, if not runnging 
root@dpu-bmc:~# systemctl enable rshim
root@dpu-bmc:~# systemctl start rshim
root@dpu-bmc:~# systemctl status rshim
root@dpu-bmc:~# ls /dev/rshim0

Repeat the step 4-10 on all your DPUs.

To authenticate with Redfish, it is necesasry to provide a password for the BMC root user.

Jump Node Console

Copy
Copied!

            
            $ kubectl create secret generic -n dpf-operator-system bmc-shared-password --from-literal=password='ROOT_BMC_PASSWORD'

Create the following YAML to detect DPUDevices and DPUNodes:

ip_discovery.yaml

Copy
Copied!

            
            ---
apiVersion: provisioning.dpu.nvidia.com/v1alpha1
kind: DPUDiscovery
metadata:
  name: dpu-discovery-192.168.1-10
  namespace: dpf-operator-system
spec:
  # Define the IP range to scan
  ipRangeSpec:
    ipRange:
      startIP: "10.0.110.201"    # Replace with your start IP
      endIP:   "10.0.110.204"     # Replace with your end IP
 
  # Optional: Set scan interval
  scanInterval: "3m"
  # Optional: Set number of workers (default is 1 per 255 IPs)
  workers: 1

Run the command to create DPUDevices and DPUNodes:

Jump Node Console

Copy
Copied!

            
            $ kubectl apply -f ip_discovery.yaml

Verify the DPF system by ensuring that the DPUDevices exist:

Jump Node Console

Copy
Copied!

            
            $ kubectl get dpudevices -n dpf-operator-system
NAME           AGE
mt2402xz0f7x   9m
mt2402xz0f80   9m
mt2402xz0f8g   9m
mt2402xz0f9n   9m

Verify the DPF system by ensuring that the DPUDevices exist.

Jump Node Console

Copy
Copied!

            
            $ kubectl get dpunodes -n dpf-operator-system
NAME                    AGE
dpu-node-mt2402xz0f7x   9m
dpu-node-mt2402xz0f80   9m
dpu-node-mt2402xz0f8g   9m
dpu-node-mt2402xz0f9n   9m

Add labels to DPUNodes. Set the values according to your environment.

Jump Node Console

Copy
Copied!

            
            kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f7x feature.node.kubernetes.io/dpu-0-pf0-name=ens1f0np0
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f7x feature.node.kubernetes.io/dpu-0-number-of-pfs=2
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f7x feature.node.kubernetes.io/dpu-oob-bridge-configured=""
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f7x feature.node.kubernetes.io/dpu-enabled=true
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f7x feature.node.kubernetes.io/dpu-0-pci-address=0000-2b-00
 
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f80 feature.node.kubernetes.io/dpu-0-pf0-name=ens1f0np0
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f80 feature.node.kubernetes.io/dpu-0-number-of-pfs=2
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f80 feature.node.kubernetes.io/dpu-oob-bridge-configured=""
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f80 feature.node.kubernetes.io/dpu-enabled=true
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f80 feature.node.kubernetes.io/dpu-0-pci-address=0000-2b-00
 
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f8g feature.node.kubernetes.io/dpu-0-pf0-name=ens1f0np0
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f8g feature.node.kubernetes.io/dpu-0-number-of-pfs=2
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f8g feature.node.kubernetes.io/dpu-oob-bridge-configured=""
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f8g feature.node.kubernetes.io/dpu-enabled=true
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f8g feature.node.kubernetes.io/dpu-0-pci-address=0000-2b-00
 
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f9n feature.node.kubernetes.io/dpu-0-pf0-name=ens1f0np0
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f9n feature.node.kubernetes.io/dpu-0-number-of-pfs=2
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f9n feature.node.kubernetes.io/dpu-oob-bridge-configured=""
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f9n feature.node.kubernetes.io/dpu-enabled=true
kubectl label dpunodes.provisioning.dpu.nvidia.com -n dpf-operator-system dpu-node-mt2402xz0f9n feature.node.kubernetes.io/dpu-0-pci-address=0000-2b-00

Use the following YAML to define a BFB resource that downloads the Bluefield Bitstream to a shared volume:

manifests/04-dpudeployment-installation/bfb.yaml

Copy
Copied!

            
            ---
apiVersion: provisioning.dpu.nvidia.com/v1alpha1
kind: BFB
metadata:
  name: bf-bundle
  namespace: dpf-operator-system
spec:
  url: $BLUEFIELD_BITSTREAM

Run the command to create the BFB:

Jump Node Console

Copy
Copied!

            
            $ cat manifests/04-dpudeployment-installation/bfb.yaml | envsubst |kubectl apply -f -