fix: update values for gpu operator network operator

[anson627]

This pull request refactors the GPU module to standardize the use of the mpi-operator namespace for MPI jobs and related resources, reorganizes configuration files, and updates code references accordingly. It also improves test coverage and correctness for MPI operator installation and configuration.

Namespace and Resource Management:

• All MPIJob resources and related Kubernetes objects are now consistently deployed in the mpi-operator namespace instead of default. This includes updating YAML manifests, code references, and resource management logic. [1] [2] [3] [4] [5] [6] [7] [8]

Configuration File Organization:

• MPI job configuration files have been moved from the cfg/mpi/ directory to cfg/mpi-operator/ to better reflect their usage and to align with the new namespace. [1] [2] [3] [4] [5]

Operator Installation and Resource Application:

• The MPI operator installation process now uses a more specific label selector and clarifies the expected chart version format.
• Network operator resource file references have been updated to use the network-operator directory, improving clarity and maintainability.

Testing Improvements:

• Unit tests for MPI operator installation and configuration have been updated to mock the correct functions and verify that the new parameters and resource paths are used, ensuring correctness after refactoring. [1] [2] [3]

Documentation:

• The README.md has been updated to remove unused configuration options for the Python script, reflecting the current set of supported arguments.

[Copilot]

Pull request overview

This PR corrects the file path locations for network operator manifest files, changing from the incorrect /net/ subdirectory to the correct /network-operator/ subdirectory. The PR also adds the corresponding configuration files that were missing.

Changes:

• Updated file paths in install_network_operator function to reference the correct network-operator subdirectory
• Added network operator configuration files (values.yaml, nfd-network-rule.yaml, nic-cluster-policy.yaml)
• Added GPU operator configuration file (values.yaml)

Reviewed changes

Copilot reviewed 1 out of 5 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
modules/python/gpu/pkg/net.py	Updated paths for NFD network rule and NIC cluster policy manifests from /net/ to /network-operator/
modules/python/gpu/cfg/network-operator/values.yaml	Added Helm values configuration for network operator with NFD node feature rules disabled
modules/python/gpu/cfg/network-operator/nic-cluster-policy.yaml	Added NIC cluster policy configuration for OFED driver and SR-IOV device plugin with InfiniBand support
modules/python/gpu/cfg/network-operator/nfd-network-rule.yaml	Added node feature rule to label nodes with Mellanox PCI devices
modules/python/gpu/cfg/gpu-operator/values.yaml	Added GPU operator Helm values with RDMA enabled and other features disabled