cloud-native-stack icon indicating copy to clipboard operation
cloud-native-stack copied to clipboard
verified publisher icon NVIDIA

Added NIC Cluster Policy CR

Open singh-kalpana opened this issue 11 months ago • 0 comments

What this PR does / why we need it?

In 24.11.2 release, Network Operator is upgraded to 24.10.1 but the corresponding updates to network-operator-values.yaml is missing. Earlier, Network Operator Helm Chart used to create the NIC Cluster policy, but this feature is removed from NetOp 24.10.0 version , requiring users to create it separately. Because of this missing step, important components like MOFED Driver, multus, whereabouts are not getting deployed. Since MOFED Driver is missing, installing the GPU Operator with RDMA causes the nvidia-driver-daemonset pod to get stuck.

Changes:

  1. Added NIC Cluster Policy CR
  2. Introduced a new variable deploy_ofed to handle the deployment of MOFED Driver via NIC Cluster Policy for NetOp 24.10.1.

singh-kalpana avatar Feb 19 '25 12:02 singh-kalpana