Fix nvidia toolkit check (#33)

* fix: nvidia toolkit check * add: Updated the nvidia driver branch * Improved container inspect --------- Co-authored-by: gioacchino.vino <[email protected]>
DODAS-TS · Jan 26, 2024 · c743f53 · c743f53
1 parent 4995ae6
commit c743f53
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/tasks/jupyterhub-partitioned-gpus.yml b/tasks/jupyterhub-partitioned-gpus.yml
@@ -45,8 +45,10 @@
   - '    running_containers_ids = client.containers()'
   - '    for id in running_containers_ids:'
   - '        my_c = client.inspect_container(id["Id"])'
-  - '        if my_c["HostConfig"]["DeviceRequests"] is not None:'
+  - '        try: '
   - '            gpu_uuid.add(my_c["HostConfig"]["DeviceRequests"][0]["DeviceIDs"][0])'
+  - '        except: '
+  - '            pass'
   - '    return gpu_uuid'
   - 'def get_available_gpu_ids():'
   - '    all_gpus_ids = get_all_gpus_ids()'

diff --git a/tasks/main.yml b/tasks/main.yml
@@ -43,8 +43,8 @@
   ansible.builtin.include_tasks: gpu-support.yaml
   vars:
     enable_gpu: true
-    gpu_support_role: git+https://baltig.infn.it/infn-cloud/ansible-role-gpu-support,v1.2.1
-    nvidia_driver_branch: "515"
+    gpu_support_role: git+https://baltig.infn.it/infn-cloud/ansible-role-gpu-support,v1.2.4
+    nvidia_driver_branch: "535"
     nvidia_driver_ubuntu_install_from_cuda_repo: true
     nvidia_driver_skip_reboot: true
   when: (use_gpu | bool)