merge conflicts

Author: wtripp180901

.github/workflows/stackhpc.yml

@@ -178,11 +178,11 @@ jobs:
           ansible-playbook -v ansible/site.yml
           ansible-playbook -v ansible/ci/check_slurm.yml
 
-      - name: Test reimage of compute nodes and compute-init (via rebuild adhoc)
+      - name: Test compute node reboot and compute-init
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
-          ansible-playbook -v --limit compute ansible/adhoc/rebuild.yml
+          ansible-playbook -v ansible/adhoc/reboot_via_slurm.yml
           ansible-playbook -v ansible/ci/check_slurm.yml
 
       - name: Check sacct state survived reimage

README.md

@@ -31,8 +31,7 @@ It requires an OpenStack cloud, and an Ansible "deploy host" with access to that
 
 Before starting ensure that:
 - You have root access on the deploy host.
-- You can create instances using a Rocky 9 GenericCloud image (or an image based on that).
-    - **NB**: In general it is recommended to use the [latest released image](https://github.com/stackhpc/ansible-slurm-appliance/releases) which already contains the required packages. This is built and tested in StackHPC's CI.
+- You can create instances from the [latest Slurm appliance image](https://github.com/stackhpc/ansible-slurm-appliance/releases), which already contains the required packages. This is built and tested in StackHPC's CI. Although you can use a Rocky Linux 9 GenericCloud instead, it is not recommended.
 - You have an SSH keypair defined in OpenStack, with the private key available on the deploy host.
 - Created instances have access to internet (note proxies can be setup through the appliance if necessary).
 - Created instances have accurate/synchronised time (for VM instances this is usually provided by the hypervisor; if not or for bare metal instances it may be necessary to configure a time service via the appliance).
@@ -82,30 +81,39 @@ And generate secrets for it:
 
 Create an OpenTofu variables file to define the required infrastructure, e.g.:
 
-    # environments/$ENV/terraform/terraform.tfvars:
+    # environments/$ENV/tofu/terraform.tfvars:
 
     cluster_name = "mycluster"
-    cluster_net = "some_network" # *
-    cluster_subnet = "some_subnet" # *
+    cluster_networks = [
+      {
+        network = "some_network" # *
+        subnet = "some_subnet" # *
+      }
+    ]
     key_pair = "my_key" # *
     control_node_flavor = "some_flavor_name"
-    login_nodes = {
-        login-0: "login_flavor_name"
+    login = {
+        # Arbitrary group name for these login nodes
+        interactive = {
+            nodes: ["login-0"]
+            flavor: "login_flavor_name" # *
+        }
     }
     cluster_image_id = "rocky_linux_9_image_uuid"
     compute = {
+        # Group name used for compute node partition definition
         general = {
             nodes: ["compute-0", "compute-1"]
-            flavor: "compute_flavor_name"
+            flavor: "compute_flavor_name" # *
         }
     }
 
-Variables marked `*` refer to OpenStack resources which must already exist. The above is a minimal configuration - for all variables and descriptions see `environments/$ENV/terraform/terraform.tfvars`.
+Variables marked `*` refer to OpenStack resources which must already exist. The above is a minimal configuration - for all variables and descriptions see `environments/$ENV/tofu/variables.tf`.
 
 To deploy this infrastructure, ensure the venv and the environment are [activated](#create-a-new-environment) and run:
 
     export OS_CLOUD=openstack
-    cd environments/$ENV/terraform/
+    cd environments/$ENV/tofu/
     tofu init
     tofu apply
 

ansible/.gitignore

@@ -32,6 +32,8 @@ roles/*
 !roles/mysql/**
 !roles/systemd/
 !roles/systemd/**
+!roles/cacerts/
+!roles/cacerts/**
 !roles/cuda/
 !roles/cuda/**
 !roles/freeipa/
@@ -82,3 +84,5 @@ roles/*
 !roles/slurm_stats/**
 !roles/pytools/
 !roles/pytools/**
+!roles/rebuild/
+!roles/rebuild/**

ansible/adhoc/reboot_via_slurm.yml

@@ -0,0 +1,24 @@
+# Reboot compute nodes via slurm. Nodes will be rebuilt if `image_id` in inventory is different to the currently-provisioned image.
+# Example:
+#   ansible-playbook -v ansible/adhoc/reboot_via_slurm.yml
+
+- hosts: login
+  run_once: true
+  become: yes
+  gather_facts: no
+  tasks:
+    - name: Submit a Slurm job to reboot compute nodes
+      ansible.builtin.shell: |
+        set -e
+        srun --reboot -N 2 uptime
+      become_user: root
+      register: slurm_result
+      failed_when: slurm_result.rc != 0
+
+    - name: Fetch Slurm controller logs if reboot fails
+      ansible.builtin.shell: |
+        journalctl -u slurmctld --since "10 minutes ago" | tail -n 50
+      become_user: root
+      register: slurm_logs
+      when: slurm_result.rc != 0
+      delegate_to: "{{ groups['control'] | first }}"

ansible/bootstrap.yml

@@ -52,6 +52,13 @@
     - import_role:
         name: proxy
 
+- hosts: chrony
+  tags: chrony
+  become: yes
+  tasks:
+    - import_role:
+        name: mrlesmithjr.chrony
+
 - hosts: cluster
   gather_facts: false
   become: yes
@@ -126,22 +133,46 @@
     ansible.builtin.assert:
       that: dnf_repos_password is undefined
       fail_msg: Passwords should not be templated into repofiles during configure, unset 'dnf_repos_password'
-    when: appliances_mode == 'configure'
-  - name: Replace system repos with pulp repos 
-    ansible.builtin.include_role:
-      name: dnf_repos
-      tasks_from: set_repos.yml
+    when:
+      - appliances_mode == 'configure'
+      - not (dnf_repos_allow_insecure_creds | default(false)) # useful for development
+
+- hosts: cacerts:!builder
+  tags: cacerts
+  gather_facts: false
+  tasks:
+    - name: Install custom cacerts
+      import_role:
+        name: cacerts
 
-# --- tasks after here require access to package repos ---
 - hosts: squid
   tags: squid
   gather_facts: yes
   become: yes
   tasks:
+    # - Installing squid requires working dnf repos
+    # - Configuring dnf_repos itself requires working dnf repos to install epel
+    # - Hence do this on squid nodes first in case they are proxying others
+    - name: Replace system repos with pulp repos
+      ansible.builtin.include_role:
+        name: dnf_repos
+        tasks_from: set_repos.yml
+      when: "'dnf_repos' in group_names"
     - name: Configure squid proxy
       import_role:
         name: squid
 
+- hosts: dnf_repos
+  tags: dnf_repos
+  gather_facts: yes
+  become: yes
+  tasks:
+  - name: Replace system repos with pulp repos 
+    ansible.builtin.include_role:
+      name: dnf_repos
+      tasks_from: set_repos.yml
+
+# --- tasks after here require general access to package repos ---
 - hosts: tuned
   tags: tuned
   gather_facts: yes
@@ -282,10 +313,11 @@
     - include_role:
         name: azimuth_cloud.image_utils.linux_ansible_init
 
-- hosts: k3s
+- hosts: k3s:&builder
   become: yes
   tags: k3s
   tasks:
-    - ansible.builtin.include_role:
+    - name: Install k3s
+      ansible.builtin.include_role:
         name: k3s
         tasks_from: install.yml

ansible/ci/check_slurm.yml

@@ -6,7 +6,7 @@
       shell: 'sinfo --noheader --format="%N %P %a %l %D %t" | sort' # using --format ensures we control whitespace: Partition,partition_state,max_jobtime,num_nodes,node_state,node_name
       register: sinfo
       changed_when: false
-      until: not ("boot" in sinfo.stdout or "idle*" in sinfo.stdout)
+      until: not ("boot" in sinfo.stdout or "idle*" in sinfo.stdout or "down" in sinfo.stdout)
       retries: 10
       delay: 5
     - name: Check nodes have expected slurm state

ansible/ci/retrieve_inventory.yml

@@ -7,7 +7,7 @@
   gather_facts: no
   vars:
     cluster_prefix: "{{ undef(hint='cluster_prefix must be defined') }}" # e.g. ci4005969475
-    ci_vars_file: "{{ appliances_environment_root + '/terraform/' + lookup('env', 'CI_CLOUD') }}.tfvars"
+    ci_vars_file: "{{ appliances_environment_root + '/tofu/' + lookup('env', 'CI_CLOUD') }}.tfvars"
     cluster_network: "{{ lookup('ansible.builtin.ini', 'cluster_net', file=ci_vars_file, type='properties') | trim('\"') }}"
   tasks:
     - name: Get control host IP

ansible/extras.yml

@@ -1,3 +1,23 @@
+- hosts: k3s_server:!builder
+  become: yes
+  tags: k3s
+  tasks:
+    - name: Start k3s server
+      ansible.builtin.include_role:
+        name: k3s
+        tasks_from: server-runtime.yml
+
+# technically should be part of bootstrap.yml but hangs waiting on failed mounts
+# if runs before filesystems.yml after the control node has been reimaged
+- hosts: k3s_agent:!builder
+  become: yes
+  tags: k3s
+  tasks:
+    - name: Start k3s agents
+      ansible.builtin.include_role:
+        name: k3s
+        tasks_from: agent-runtime.yml
+
 - hosts: basic_users:!builder
   become: yes
   tags:

ansible/roles/basic_users/README.md

@@ -2,16 +2,19 @@
 basic_users
 ===========
 
-Setup users on cluster nodes using `/etc/passwd` and manipulating `$HOME`, i.e. without requiring LDAP etc. Features:
+Setup users on cluster nodes using `/etc/passwd` and manipulating `$HOME`, i.e.
+without requiring LDAP etc. Features:
 - UID/GID is consistent across cluster (and explicitly defined).
 - SSH key generated and propagated to all nodes to allow login between cluster nodes.
 - An "external" SSH key can be added to allow login from elsewhere.
-- Login to the control node is prevented.
+- Login to the control node is prevented (by default)
 - When deleting users, systemd user sessions are terminated first.
 
 Requirements
 ------------
-- $HOME (for normal users, i.e. not `centos`) is assumed to be on a shared filesystem.
+- `$HOME` (for normal users, i.e. not `rocky`) is assumed to be on a shared
+  filesystem. Actions affecting that shared filesystem are run on a single host,
+  see `basic_users_manage_homedir` below.
 
 Role Variables
 --------------
@@ -22,9 +25,15 @@ Role Variables
   - `shell` if *not* set will be `/sbin/nologin` on the `control` node and the default shell on other users. Explicitly setting this defines the shell for all nodes.
   - An additional key `public_key` may optionally be specified to define a key to log into the cluster.
   - An additional key `sudo` may optionally be specified giving a string (possibly multiline) defining sudo rules to be templated.
+  - `ssh_key_type` defaults to `ed25519` instead of the `ansible.builtin.user` default of `rsa`.
   - Any other keys may present for other purposes (i.e. not used by this role).
 - `basic_users_groups`: Optional, default empty list. A list of mappings defining information for each group. Mapping keys/values are passed through as parameters to [ansible.builtin.group](https://docs.ansible.com/ansible/latest/collections/ansible/builtin/group_module.html) and default values are as given there.
 - `basic_users_override_sssd`: Optional bool, default false. Whether to disable `sssd` when ensuring users/groups exist with this role. Permits creating local users/groups even if they clash with users provided via sssd (e.g. from LDAP). Ignored if host is not in group `sssd` as well. Note with this option active `sssd` will be stopped and restarted each time this role is run.
+- `basic_users_manage_homedir`: Optional bool, must be true on a single host to
+  determine which host runs tasks affecting the shared filesystem. The default
+  is to use the first play host which is not the control node, because the
+  default NFS configuration does not have the shared `/home` directory mounted
+  on the control node.
 
 Dependencies
 ------------

ansible/roles/basic_users/defaults/main.yml

@@ -1,9 +1,10 @@
-basic_users_manage_homedir: "{{ (ansible_hostname == (ansible_play_hosts | first)) }}"
+basic_users_manage_homedir: "{{ ansible_hostname == (ansible_play_hosts | difference(groups['control']) | first) }}"
 basic_users_userdefaults:
   state: present
   create_home: "{{ basic_users_manage_homedir }}"
   generate_ssh_key:  "{{ basic_users_manage_homedir }}"
   ssh_key_comment: "{{ item.name }}"
+  ssh_key_type: ed25519
   shell: "{{'/sbin/nologin' if 'control' in group_names else omit }}"
 basic_users_users: []
 basic_users_groups: []