Merge pull request #18 from svc-design/codex/fix-permission-issue-with-get_labring_registry.sh

Fix GPU role variable checks

docs/gpu-k8s-role.md

@@ -59,8 +59,26 @@ Example playbook snippet defining the IP lists:
     - gpu-k8s
 ```
 
-The playbook expects `master_ips` and `node_ips` variables which are lists of IP addresses. Up to
-three masters can be specified.
+You can also specify hostnames and let the role look up the IPs:
+
+```yaml
+- hosts: all
+  vars:
+    masters:
+      - "k8s-1"
+    nodes:
+      - "k8s-2"
+      - "k8s-3"
+  roles:
+    - ssh-trust
+    - gpu-k8s
+```
+
+The playbook expects at least one master and one node. You can provide the
+addresses directly via `master_ips` and `node_ips`, or give hostnames in the
+`masters` and `nodes` variables. When hostnames are used, the role will look up
+their `ansible_host` values from the inventory to obtain the IPs. Up to three
+masters can be specified.
 
 
 Run the playbook with your inventory that contains the master and node IP addresses.

playbooks/roles/vhosts/gpu-k8s/defaults/main.yml

@@ -3,6 +3,8 @@ sealos_version: latest
 kubernetes_version: v1.25.16
 cilium_version: v1.13.4
 helm_version: v3.9.4
+masters: []    # List of master hostnames
+nodes: []      # List of worker hostnames
 master_ips: []  # List of up to three master node IPs
 node_ips: []    # List of worker node IPs
 sealos_cmd_env: '{}'

playbooks/roles/vhosts/gpu-k8s/tasks/install_cluster.yml

@@ -10,6 +10,18 @@
     sealos_version: "{{ (sealos_latest.content | from_json).tag_name }}"
   when: sealos_version is not defined or sealos_version == 'latest'
 
+- name: Resolve master and node IPs from hostnames when needed
+  set_fact:
+    master_ips: >-
+      {{ (master_ips | default([])) | length > 0
+         | ternary(master_ips,
+                   masters | default([]) | map('extract', hostvars, 'ansible_host') | list) }}
+    node_ips: >-
+      {{ (node_ips | default([])) | length > 0
+         | ternary(node_ips,
+                   nodes | default([]) | map('extract', hostvars, 'ansible_host') | list) }}
+
+
 - name: Install sealos CLI
   shell: |
     VERSION={{ sealos_version }}
@@ -19,7 +31,7 @@
     mv sealos /usr/bin/sealos
   args:
     executable: /bin/bash
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Install Helm
   shell: |
@@ -29,7 +41,7 @@
     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')
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first)) and (ansible_facts['distribution'] == 'Ubuntu' or ansible_facts['distribution'] == 'Debian')
 
 - name: Install nerdctl
   shell: |
@@ -37,7 +49,7 @@
     tar -C /usr/local -xzf nerdctl-2.0.2-linux-amd64.tar.gz
   args:
     executable: /bin/bash
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Determine LabRing registry prefix
   shell: "{{ role_path }}/files/get_labring_registry.sh"
@@ -46,7 +58,15 @@
   delegate_to: localhost
   run_once: true
   become: false
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
+
+- name: Validate master_ips and node_ips
+  assert:
+    that:
+      - master_ips | length > 0
+      - node_ips | length > 0
+    fail_msg: "Provide masters/nodes hostnames or master_ips/node_ips with at least one entry"
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Run sealos to create Kubernetes cluster
   shell: |
@@ -60,13 +80,13 @@
       --cmd "{{ kubeadm_init_cmd }}"
   args:
     executable: /bin/bash
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Show sealos CLI version
   command: sealos version
   register: sealos_cli_version
   changed_when: false
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Display Kubernetes cluster status
   shell: kubectl get nodes -o wide
@@ -74,9 +94,9 @@
     executable: /bin/bash
   register: k8s_status
   changed_when: false
-  when: inventory_hostname == (ops_host | default(master_ips | first))
+  when: inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Print cluster nodes
   debug:
     msg: "{{ k8s_status.stdout }}"
-  when: k8s_status is defined and inventory_hostname == (ops_host | default(master_ips | first))
+  when: k8s_status is defined and inventory_hostname == (ops_host | default(masters | default(master_ips) | first))

playbooks/roles/vhosts/gpu-k8s/tasks/run_test.yml

@@ -2,7 +2,7 @@
   shell: kubectl apply -f https://raw.githubusercontent.com/NVIDIA/k8s-device-plugin/v0.14.5/nvidia-device-plugin.yml
   args:
     executable: /bin/bash
-  when: master_ips|length > 0 and inventory_hostname == (master_ips | first)
+  when: master_ips|length > 0 and inventory_hostname == (ops_host | default(masters | default(master_ips) | first))
 
 - name: Run CUDA validation pod
   shell: |
@@ -10,5 +10,5 @@
     kubectl delete pod gpu-test --wait
   args:
     executable: /bin/bash
-  when: master_ips|length > 0 and inventory_hostname == (master_ips | first)
+  when: master_ips|length > 0 and inventory_hostname == (ops_host | default(masters | default(master_ips) | first))