Merge branch 'main' into fix/lustre-lu-18085

Author: sjpb

ansible/roles/cuda/README.md

@@ -10,6 +10,6 @@ Requires OFED to be installed to provide required kernel-* packages.
 
 - `cuda_repo_url`: Optional. URL of `.repo` file. Default is upstream for appropriate OS/architecture.
 - `cuda_nvidia_driver_stream`: Optional. Version of `nvidia-driver` stream to enable. This controls whether the open or proprietary drivers are installed and the major version. Changing this once the drivers are installed does not change the version.
-- `cuda_packages`: Optional. Default: `['cuda', 'nvidia-gds', 'cmake', 'cuda-toolkit-12-8']`.
+- `cuda_packages`: Optional. Default: `['cuda', 'nvidia-gds', 'cmake', 'cuda-toolkit-12-9']`.
 - `cuda_package_version`: Optional. Default `latest` which will install the latest packages if not installed but won't upgrade already-installed packages. Use `'none'` to skip installing CUDA.
 - `cuda_persistenced_state`: Optional. State of systemd `nvidia-persistenced` service. Values as [ansible.builtin.systemd:state](https://docs.ansible.com/ansible/latest/collections/ansible/builtin/systemd_module.html#parameter-state). Default `started`.

ansible/roles/cuda/defaults/main.yml

@@ -1,12 +1,12 @@
 cuda_repo_url: "https://developer.download.nvidia.com/compute/cuda/repos/rhel{{ ansible_distribution_major_version }}/{{ ansible_architecture }}/cuda-rhel{{ ansible_distribution_major_version }}.repo"
-cuda_nvidia_driver_stream: '570-open'
-cuda_package_version: '12.8.1-1'
-cuda_version_short: '12.8'
+cuda_nvidia_driver_stream: '575-open'
+cuda_package_version: '12.9.0-1'
+cuda_version_short: '12.9'
 cuda_packages:
   - "cuda{{ ('-' + cuda_package_version) if cuda_package_version != 'latest' else '' }}"
   - nvidia-gds
   - cmake
-  - cuda-toolkit-12-8
+  - cuda-toolkit-12-9
 cuda_samples_release_url: "https://github.com/NVIDIA/cuda-samples/archive/refs/tags/v{{ cuda_version_short }}.tar.gz"
 cuda_samples_path: "/var/lib/{{ ansible_user }}/cuda_samples"
 cuda_samples_programs:

ansible/roles/cuda/tasks/samples.yml

@@ -25,36 +25,3 @@
     cmd: . /etc/profile.d/sh.local && cmake .. && make -j {{ ansible_processor_vcpus }}
     chdir: "{{ cuda_samples_path }}/cuda-samples-{{ cuda_version_short }}/build"
     creates: "{{ cuda_samples_path }}/cuda-samples-{{ cuda_version_short }}/build/Samples/1_Utilities/deviceQuery/deviceQuery"
-
-- name: Run CUDA deviceQuery
-  command:
-    cmd: "{{ cuda_samples_path }}/cuda-samples-{{ cuda_version_short }}/build/Samples/1_Utilities/deviceQuery/deviceQuery"
-  register: _cuda_devicequery
-
-- name: Set fact for CUDA devices
-  set_fact:
-    cuda_devices: "{{ _cuda_devicequery.stdout | regex_findall('Device (\\d+):') }}"
-
-- name: Run CUDA bandwidth test
-  command:
-    cmd: "{{ cuda_samples_path }}/cuda-samples-{{ cuda_version_short }}/build/Samples/1_Utilities/bandwidthTest/bandwidthTest --device={{ item }}"
-  register: _cuda_bandwidthtest
-  loop: "{{ cuda_devices }}"
-  loop_control:
-    label: "Device {{ item }}" # e.g '0'
-
-- name: Summarise bandwidth test output
-  debug:
-    msg: |
-      {{ _parts[1].splitlines()[0] | trim }}
-      Bandwidths: (Gb/s)
-      Host to Device: {{ _parts[2].split()[-1] }}
-      Device to Host: {{ _parts[3].split()[-1] }}
-      Device to Device: {{ _parts[4].split()[-1] }}
-      {{ ': '.join(_parts[5].split('=') | map('trim')) }}
-      {{ _parts[6] }}
-  loop: "{{ _cuda_bandwidthtest.results }}"
-  vars:
-    _parts: "{{ item.stdout.split('\n\n') }}"
-  loop_control:
-    label: "Device {{ item.item }}" # e.g '0'