ai-workspace-infra/gitops
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

gpu role: fetch latest sealos and install tools

Browse Source
This commit is contained in:
shenlan 2025-06-25 21:33:22 +08:00
parent 9d19914dec
commit 3933de1764
3 changed files with 42 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
docs/gpu-k8s-role.md
View File

@ -11,14 +11,17 @@ The role performs four main tasks:
3. **Verify the cluster state** after initialization, displaying the `sealos` version and the current Kubernetes nodes. 3. **Verify the cluster state** after initialization, displaying the `sealos` version and the current Kubernetes nodes.
4. **Verify GPU access** by deploying the official NVIDIA device plugin and running a small CUDA workload. 4. **Verify GPU access** by deploying the official NVIDIA device plugin and running a small CUDA workload.
When `sealos_version` is set to `latest` (the default), the role automatically
fetches the most recent stable release from GitHub.
The following command is used to create the cluster (example with one master and one worker): The following command is used to create the cluster (example with one master and one worker):
```bash ```bash
sealos run \ sealos run \
registry.cn-shanghai.aliyuncs.com/labring/kubernetes:v1.29.9 \ registry.cn-shanghai.aliyuncs.com/labring/kubernetes:<latest> \
registry.cn-shanghai.aliyuncs.com/labring/cilium:v1.13.4 \ registry.cn-shanghai.aliyuncs.com/labring/cilium:<cilium_version> \
registry.cn-shanghai.aliyuncs.com/labring/helm:v3.9.4 \ registry.cn-shanghai.aliyuncs.com/labring/helm:<helm_version> \
--masters 172.16.11.120 \ --masters 172.16.11.120 \
--nodes 172.16.11.152 \ --nodes 172.16.11.152 \
--env '{}' \ --env '{}' \

2
playbooks/roles/vhosts/gpu-k8s/defaults/main.yml
View File

@ -1,5 +1,5 @@
# Default variables for gpu-k8s role # Default variables for gpu-k8s role
sealos_version: v1.29.9 sealos_version: latest
cilium_version: v1.13.4 cilium_version: v1.13.4
helm_version: v3.9.4 helm_version: v3.9.4
master_ips: [] # List of up to three master node IPs master_ips: [] # List of up to three master node IPs

36
playbooks/roles/vhosts/gpu-k8s/tasks/install_cluster.yml
View File

@ -1,6 +1,40 @@
- name: Determine latest sealos version when requested
uri:
url: https://api.github.com/repos/labring/sealos/releases/latest
return_content: yes
register: sealos_latest
when: sealos_version is not defined or sealos_version == 'latest'
- name: Set sealos_version fact to latest release
set_fact:
sealos_version: "{{ (sealos_latest.content | from_json).tag_name }}"
when: sealos_version is not defined or sealos_version == 'latest'
- name: Install sealos CLI - name: Install sealos CLI
shell: | shell: |
curl -sL https://raw.githubusercontent.com/labring/sealos/main/scripts/install.sh | bash -s {{ sealos_version }} VERSION={{ sealos_version }}
wget -q https://github.com/labring/sealos/releases/download/${VERSION}/sealos_${VERSION#v}_linux_amd64.tar.gz
tar -xzf sealos_${VERSION#v}_linux_amd64.tar.gz sealos
chmod +x sealos
mv sealos /usr/bin/sealos
args:
executable: /bin/bash
when: inventory_hostname == (ops_host | default(master_ips | first))
- name: Install Helm
shell: |
curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3
chmod 700 get_helm.sh
./get_helm.sh
rm -f get_helm.sh
args:
executable: /bin/bash
when: inventory_hostname == (ops_host | default(master_ips | first)) and (ansible_facts['distribution'] == 'Ubuntu' or ansible_facts['distribution'] == 'Debian')
- name: Install nerdctl
shell: |
wget -q https://github.com/containerd/nerdctl/releases/download/v2.0.2/nerdctl-2.0.2-linux-amd64.tar.gz
tar -C /usr/local -xzf nerdctl-2.0.2-linux-amd64.tar.gz
args: args:
executable: /bin/bash executable: /bin/bash
when: inventory_hostname == (ops_host | default(master_ips | first)) when: inventory_hostname == (ops_host | default(master_ips | first))