Merge branch 'feat/alertmanager' of github.com:stackhpc/ansible-slurm-appliance into feat/alertmanager

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']`.
+- `cuda_packages`: Optional. Default: `['cuda', 'nvidia-gds', 'cmake', 'cuda-toolkit-12-8']`.
 - `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

@@ -1,13 +1,14 @@
 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_packages:
   - "cuda{{ ('-' + cuda_package_version) if cuda_package_version != 'latest' else '' }}"
   - nvidia-gds
-# _cuda_version_tuple: # discovered from installed package e.g. ('12', '1', '0')
-cuda_version_short: "{{ _cuda_version_tuple[0] }}.{{ _cuda_version_tuple[1] }}"
+  - cmake
+  - cuda-toolkit-12-8
 cuda_samples_release_url: "https://github.com/NVIDIA/cuda-samples/archive/refs/tags/v{{ cuda_version_short }}.tar.gz"
-cuda_samples_path: "/home/{{ ansible_user }}/cuda_samples"
+cuda_samples_path: "/var/lib/{{ ansible_user }}/cuda_samples"
 cuda_samples_programs:
   - deviceQuery
   - bandwidthTest

ansible/roles/cuda/tasks/samples.yml

@@ -1,12 +1,3 @@
-- name: Read CUDA version file
-  slurp:
-    src: /usr/local/cuda/version.json
-  register: _cuda_samples_version
-
-- name: Set fact for discovered CUDA version
-  set_fact:
-    _cuda_version_tuple: "{{ (_cuda_samples_version.content | b64decode | from_json).cuda.version | split('.') }}" # e.g. '12.1.0'
-
 - name: Ensure cuda_samples_path exists
   file:
     state: directory

ansible/roles/eessi/defaults/main.yaml

@@ -9,3 +9,5 @@ cvmfs_config_default:
 cvmfs_config_overrides: {}
 
 cvmfs_config: "{{ cvmfs_config_default | combine(cvmfs_config_overrides) }}"
+
+cvmfs_gpg_checksum: "sha256:4ac81adff957565277cfa6a4a330cdc2ce5a8fdd73b8760d1a5a32bef71c4bd6"

ansible/roles/eessi/tasks/main.yaml

@@ -1,8 +1,9 @@
 ---
 - name: Download Cern GPG key
   ansible.builtin.get_url:
-    url: http://cvmrepo.web.cern.ch/cvmrepo/yum/RPM-GPG-KEY-CernVM 
+    url: http://cvmrepo.web.cern.ch/cvmrepo/yum/RPM-GPG-KEY-CernVM
     dest: ./cvmfs-key.gpg
+    checksum: "{{ cvmfs_gpg_checksum }}"
 
 - name: Import downloaded GPG key
   command: rpm --import cvmfs-key.gpg
@@ -24,7 +25,7 @@
 
 # Alternative version using official repo - still no GPG key :(
 # - name: Add EESSI repo
-#   dnf: 
+#   dnf:
 #     name: http://repo.eessi-infra.org/eessi/rhel/8/noarch/eessi-release-0-1.noarch.rpm
 
 # - name: Install EESSI CVMFS config
@@ -39,9 +40,9 @@
     value: "{{ item.value }}"
     no_extra_spaces: true
   loop: "{{ cvmfs_config | dict2items }}"
-    
+
 
 # NOTE: Not clear how to make this idempotent
 - name: Ensure CVMFS config is setup
   command:
-    cmd: "cvmfs_config setup"
+    cmd: "cvmfs_config setup"