Adds bandwidth.yml playbook for NVIDIA nvbandwidth and

removes samples.yml tasks from adhoc/cudatest.yml

Author: claudia-lola

ansible/adhoc/cudatests.yml

@@ -2,12 +2,8 @@
 - hosts: cuda
   become: true
   gather_facts: true
-  tags: cuda_samples
+  tags: cuda_bandwidth
   tasks:
-    - ansible.builtin.import_role:
-        name: cuda
-        tasks_from: samples.yml
-
     - name: Run CUDA bandwidth tasks
       ansible.builtin.import_role:
         name: cuda

ansible/roles/cuda/defaults/main.yml

@@ -16,6 +16,8 @@ cuda_samples_programs:
   - bandwidthTest
 # cuda_devices: # discovered from deviceQuery run
 cuda_persistenced_state: started
+cuda_install_nvidiafabricmanger: false
 # variables for nvbandwidth (for bandwidth.yml tasks run in cudatests.yml)
+cuda_bandwidth_version: '0.8'
 cuda_bandwidth_path: "/var/lib/{{ ansible_user }}/cuda_bandwidth"
-cuda_bandwidth_release_url: "https://github.com/NVIDIA/nvbandwidth/archive/refs/tags/v0.8.tar.gz"
+cuda_bandwidth_release_url: "https://github.com/NVIDIA/nvbandwidth/archive/refs/tags/v{{ cuda_bandwidth_version }}.tar.gz"

ansible/roles/cuda/tasks/bandwidth.yml

@@ -1,5 +1,5 @@
 ---
-- name: Ensure cuda_bandwidth_path exists
+- name: Ensure CUDA bandwidth path exists
   ansible.builtin.file:
     state: directory
     path: "{{ cuda_bandwidth_path }}"
@@ -14,12 +14,12 @@
     dest: "{{ cuda_bandwidth_path }}"
     owner: "{{ ansible_user }}"
     group: "{{ ansible_user }}"
-    creates: "{{ cuda_bandwidth_path }}/nvbandwidth-0.8"
+    creates: "{{ cuda_bandwidth_path }}/nvbandwidth-{{ cuda_bandwidth_version }}"
 
 - name: Creates CUDA bandwidth test build directory
   ansible.builtin.file:
     state: directory
-    path: "{{ cuda_bandwidth_path }}/nvbandwidth-0.8/build"
+    path: "{{ cuda_bandwidth_path }}/nvbandwidth-{{ cuda_bandwidth_version }}/build"
     mode: "0755"
 
 - name: Ensure cudatests directory exists
@@ -31,26 +31,27 @@
 
 - name: Build CUDA bandwidth test
   ansible.builtin.shell:
-    cmd: |
+    cmd: >
       source /cvmfs/software.eessi.io/versions/2023.06/init/bash &&
+      module load buildenv/default-foss-2023b &&
       module load Boost/1.82.0-GCC-12.3.0 &&
-      . /etc/profile.d/sh.local && cmake .. &&
+      . /etc/profile.d/sh.local &&
+      cmake .. &&
       make -j {{ ansible_processor_vcpus }}
-    chdir: "{{ cuda_bandwidth_path }}/nvbandwidth-0.8/build"
-    creates: "{{ cuda_bandwidth_path }}/nvbandwidth-0.8/build/nvbandwidth"
+    chdir: "{{ cuda_bandwidth_path }}/nvbandwidth-{{ cuda_bandwidth_version }}/build"
+    creates: "{{ cuda_bandwidth_path }}/nvbandwidth-{{ cuda_bandwidth_version }}/build/nvbandwidth"
 
 - name: Run CUDA bandwidth test
   ansible.builtin.shell: |
-    export LD_LIBRARY_PATH=/cvmfs/software.eessi.io/versions/2023.06/software/linux/x86_64/amd/zen4/software/GCCcore/12.3.0/lib64:\
-    /cvmfs/software.eessi.io/versions/2023.06/software/linux/x86_64/amd/zen4/software/Boost/1.82.0-GCC-12.3.0/lib
     ./nvbandwidth
   args:
-    chdir: "{{ cuda_bandwidth_path }}/nvbandwidth-0.8/build/"
+    chdir: "{{ cuda_bandwidth_path }}/nvbandwidth-{{ cuda_bandwidth_version }}/build/"
   register: cuda_bandwidth_output
+  changed_when: true
 
 - name: Save CUDA bandwidth output to bandwidth_results.txt
   ansible.builtin.copy:
     content: "{{ cuda_bandwidth_output.stdout }}"
-    dest: "{{ appliances_environment_root }}/cudatests/bandwidth_results.txt"
+    dest: "{{ appliances_environment_root }}/cudatests/{{ inventory_hostname }}bandwidth_results.txt"
     mode: '0644'
   delegate_to: localhost