initial pfannkuchen

roles/nvidia/defaults/main.yml

@@ -0,0 +1,8 @@
+---
+cuda_keyring_deb_url: "https://developer.download.nvidia.com/compute/cuda/repos/debian13/x86_64/cuda-keyring_1.1-1_all.deb"
+docker_daemon_config:
+  default-runtime: nvidia
+  runtimes:
+    nvidia:
+      path: nvidia-container-runtime
+      runtimeArgs: []

roles/nvidia/handlers/main.yml

@@ -0,0 +1,5 @@
+---
+- name: Docker neustarten
+  ansible.builtin.service:
+    name: docker
+    state: restarted

roles/nvidia/tasks/main.yml

@@ -0,0 +1,77 @@
+---
+- name: Nouveau Treiber blacklisten
+  ansible.builtin.copy:
+    dest: /etc/modprobe.d/blacklist-nouveau.conf
+    content: |
+      blacklist nouveau
+      options nouveau modeset=0
+    mode: "0644"
+  register: nouveau_blacklist
+
+- name: initramfs-tools installieren
+  apt:
+    name: initramfs-tools
+    state: present
+
+- name: initramfs aktualisieren
+  ansible.builtin.command: update-initramfs -u
+  when: nouveau_blacklist.changed
+
+- name: CUDA Keyring herunterladen und installieren
+  ansible.builtin.apt:
+    deb: "{{ cuda_keyring_deb_url }}"
+
+- name: NVIDIA Container Toolkit GPG Key hinzufuegen
+  ansible.builtin.shell: |
+    curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey \
+      | gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
+  args:
+    creates: /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
+
+- name: NVIDIA Container Toolkit Repo hinzufuegen
+  ansible.builtin.shell: |
+    curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list \
+      | sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' \
+      | tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
+  args:
+    creates: /etc/apt/sources.list.d/nvidia-container-toolkit.list
+
+- name: apt update
+  ansible.builtin.apt:
+    update_cache: true
+
+- name: Kernel Headers und DKMS installieren
+  ansible.builtin.apt:
+    name:
+      - linux-headers-{{ ansible_kernel }}
+      - dkms
+    state: present
+
+- name: NVIDIA Treiber installieren (CUDA Repo)
+  ansible.builtin.apt:
+    name: cuda-drivers
+    state: present
+  register: nvidia_driver
+
+- name: NVIDIA Container Toolkit installieren
+  ansible.builtin.apt:
+    name:
+      - nvidia-container-toolkit
+    state: present
+
+- name: Docker daemon.json konfigurieren (NVIDIA als default runtime)
+  ansible.builtin.copy:
+    content: "{{ docker_daemon_config | to_nice_json }}"
+    dest: /etc/docker/daemon.json
+    mode: "0644"
+  notify: Docker neustarten
+
+- name: nvidia-ctk runtime fuer Docker konfigurieren
+  ansible.builtin.command: nvidia-ctk runtime configure --runtime=docker
+  changed_when: false
+
+- name: Reboot nach Treiberinstallation
+  ansible.builtin.reboot:
+    msg: "Reboot nach NVIDIA Treiber-Installation"
+    reboot_timeout: 300
+  when: nvidia_driver.changed