Update appliance for stackhpc.openhpc nodegroup/partition changes

ansible/roles/cluster_infra/templates/outputs.tf.j2

@@ -32,12 +32,12 @@ output "cluster_nodes" {
         }
       }
     ],
-    {% for partition in openhpc_slurm_partitions %}
+    {% for nodegroup in openhpc_nodegroups %}
     [
-      for compute in openstack_compute_instance_v2.{{ partition.name }}: {
+      for compute in openstack_compute_instance_v2.{{ nodegroup.name }}: {
         name          = compute.name
         ip            = compute.network[0].fixed_ip_v4
-        groups        = ["compute", "{{ cluster_name }}_compute", "{{ cluster_name }}_{{ partition.name }}"],
+        groups        = ["compute", "{{ cluster_name }}_compute", "{{ cluster_name }}_{{ nodegroup.name }}"],
         facts  = {
           openstack_project_id = data.openstack_identity_auth_scope_v3.scope.project_id
         }

ansible/roles/cluster_infra/templates/resources.tf.j2

@@ -282,11 +282,11 @@ resource "openstack_networking_port_v2" "control_storage" {
 ###
 # Workers
 ###
-{% for partition in openhpc_slurm_partitions %}
+{% for nodegroup in openhpc_nodegroups %}
 # Primary network
-resource "openstack_networking_port_v2" "{{ partition.name }}" {
-  count = {{ partition.count }}
-  name = "{{ cluster_name }}-compute-{{ partition.name }}-${count.index}"
+resource "openstack_networking_port_v2" "{{ nodegroup.name }}" {
+  count = {{ nodegroup.count }}
+  name = "{{ cluster_name }}-compute-{{ nodegroup.name }}-${count.index}"
   network_id = "${data.openstack_networking_network_v2.cluster_network.id}"
   admin_state_up = "true"
 
@@ -305,9 +305,9 @@ resource "openstack_networking_port_v2" "{{ partition.name }}" {
 
 # Storage network
 {% if cluster_storage_network is defined %}
-resource "openstack_networking_port_v2" "{{ partition.name }}_storage" {
-  count          = {{ partition.count }}
-  name           = "{{ cluster_name }}-compute-{{ partition.name }}-storage-${count.index}"
+resource "openstack_networking_port_v2" "{{ nodegroup.name }}_storage" {
+  count          = {{ nodegroup.count }}
+  name           = "{{ cluster_name }}-compute-{{ nodegroup.name }}-storage-${count.index}"
   network_id     = data.openstack_networking_network_v2.cluster_storage.id
   admin_state_up = "true"
 
@@ -499,25 +499,25 @@ resource "openstack_compute_instance_v2" "control" {
   }
 }
 
-{% for partition in openhpc_slurm_partitions %}
-resource "openstack_compute_instance_v2" "{{ partition.name }}" {
-  count = {{ partition.count }}
+{% for nodegroup in openhpc_nodegroups %}
+resource "openstack_compute_instance_v2" "{{ nodegroup.name }}" {
+  count = {{ nodegroup.count }}
 
-  name      = "{{ cluster_name }}-compute-{{ partition.name }}-${count.index}"
+  name      = "{{ cluster_name }}-compute-{{ nodegroup.name }}-${count.index}"
   image_id  = "{{ cluster_image }}"
-  {% if 'flavor_name' in partition %}
-  flavor_name = "{{ partition.flavor_name }}"
+  {% if 'flavor_name' in nodegroup %}
+  flavor_name = "{{ nodegroup.flavor_name }}"
   {% else %}
-  flavor_id = "{{ partition.flavor }}"
+  flavor_id = "{{ nodegroup.flavor }}"
   {% endif %}
 
   network {
-    port = openstack_networking_port_v2.{{ partition.name }}[count.index].id
+    port = openstack_networking_port_v2.{{ nodegroup.name }}[count.index].id
   }
 
   {% if cluster_storage_network is defined %}
   network {
-    port = openstack_networking_port_v2.{{ partition.name }}_storage[count.index].id
+    port = openstack_networking_port_v2.{{ nodegroup.name }}_storage[count.index].id
   }
   {% endif %}
 

ansible/roles/openondemand/README.md

@@ -59,10 +59,10 @@ This role enables SSL on the Open Ondemand server, using the following self-sign
     - `new_window`: Optional. Whether to open link in new window. Bool, default `false`.
     - `app_name`: Optional. Unique name for app appended to `/var/www/ood/apps/sys/`. Default is `name`, useful if that is not unique or not suitable as a path component.
 - `openondemand_dashboard_support_url`: Optional. URL or email etc to show as support contact under Help in dashboard. Default `(undefined)`.
-- `openondemand_desktop_partition`: Optional. Name of Slurm partition to use for remote desktops. Requires a corresponding group named "openondemand_desktop" and entry in openhpc_slurm_partitions.
+- `openondemand_desktop_partition`: Optional. Name of Slurm partition to use for remote desktops. Requires a corresponding group named "openondemand_desktop" and entry in openhpc_partitions.
 - `openondemand_desktop_screensaver`: Optional. Whether to enable screen locking/screensaver. **NB:** Users must have passwords if this is enabled. Bool, default `false`.
 - `openondemand_filesapp_paths`: List of paths (in addition to $HOME, which is always added) to include shortcuts to within the Files dashboard app.
-- `openondemand_jupyter_partition`: Required. Name of Slurm partition to use for Jupyter Notebook servers. Requires a corresponding group named "openondemand_jupyter" and entry in openhpc_slurm_partitions.
+- `openondemand_jupyter_partition`: Required. Name of Slurm partition to use for Jupyter Notebook servers. Requires a corresponding group named "openondemand_jupyter" and entry in openhpc_partitions.
 
 ### Monitoring
 - `openondemand_exporter`: Optional. Install the Prometheus [ondemand_exporter](https://github.com/OSC/ondemand_exporter) on the `openondemand` node to export metrics about Open Ondemand itself. Default `true`.

ansible/roles/rebuild/README.md

@@ -23,7 +23,7 @@ The below are only used by this role's `rebuild.yml` task file, i.e. when
 running the `ansible/adhoc/rebuild-via-slurm.yml` playbook:
 
 - `rebuild_job_partitions`: Optional. Comma-separated list of names of rebuild
-  partitions defined in `openhpc_slurm_partitions`. Useful as an extra-var for
+  partitions defined in `openhpc_partitions`. Useful as an extra-var for
   limiting rebuilds. Default `rebuild`.
 
 - `rebuild_job_name`: Optional. Name of rebuild jobs. Default is `rebuild-`

ansible/validate.yml

@@ -23,12 +23,17 @@
   gather_facts: false
   tags: openhpc
   tasks:
+    - import_role:
+        name: stackhpc.openhpc
+        tasks_from: validate.yml
     - assert:
         that: "'enable_configless' in openhpc_config.SlurmctldParameters | default([])"
         fail_msg: |
           'enable_configless' not found in openhpc_config.SlurmctldParameters - is variable openhpc_config overridden?
           Additional slurm.conf parameters should be provided using variable openhpc_config_extra.
         success_msg: Checked Slurm will be configured for configless operation
+      delegate_to: localhost
+      run_once: true
 
 - name: Validate filebeat configuration
   hosts: filebeat

docs/experimental/slurm-controlled-rebuild.md

@@ -107,42 +107,17 @@ The configuration of this is complex and involves:
       defined in the `compute` or `login` variables, to override the default
       image for specific node groups.
 
-5. Modify `openhpc_slurm_partitions` to add a new partition covering rebuildable
-   nodes to use for for rebuild jobs. If using the default OpenTofu
-   configurations, this variable is contained in an OpenTofu-templated file
-   `environments/$ENV/group_vars/all/partitions.yml` which must be overriden
-   by copying it to e.g. a `z_partitions.yml` file in the same directory.
-   However production sites will probably be overriding this file anyway to
-   customise it.
-
-   An example partition definition, given the two node groups "general" and
-   "gpu" shown in Step 2, is:
-
-    ```yaml
-    openhpc_slurm_partitions:
-        ...
-        - name: rebuild
-          groups:
-            - name: general
-            - name: gpu
-          default: NO
-          maxtime: 30
-          partition_params:
-            PriorityJobFactor: 65533
-            Hidden: YES
-            RootOnly: YES
-            DisableRootJobs: NO
-            PreemptMode: 'OFF'
-            OverSubscribe: EXCLUSIVE
-    ```
-
-    Which has parameters as follows:
+5. Ensure `openhpc_partitions` contains a partition covering the nodes to run
+   rebuild jobs. The default definition in `environments/common/inventory/group_vars/all/openhpc.yml`
+   will automatically include this via `openhpc_rebuild_partition` also in that
+   file. If modifying this, note the important parameters are:
+
     - `name`: Partition name matching `rebuild` role variable `rebuild_partitions`,
       default `rebuild`.
-    - `groups`: A list of node group names, matching keys in the OpenTofu
-      `compute` variable (see example in step 2 above). Normally every compute
-      node group should be listed here, unless Slurm-controlled rebuild is not
-      required for certain node groups.
+    - `groups`: A list of nodegroup names, matching `openhpc_nodegroup` and
+      keys in the OpenTofu `compute` variable (see example in step 2 above).
+      Normally every compute node group should be listed here, unless
+      Slurm-controlled rebuild is not required for certain node groups.
     - `default`: Must be set to `NO` so that it is not the default partition.
     - `maxtime`: Maximum time to allow for rebuild jobs, in
       [slurm.conf format](https://slurm.schedmd.com/slurm.conf.html#OPT_MaxTime).

environments/.caas/inventory/group_vars/all/openhpc.yml

@@ -1,5 +1,5 @@
 openhpc_cluster_name: "{{ cluster_name }}"
 
-# Provision a single "standard" compute partition using the supplied
+# Provision a single "standard" compute nodegroup using the supplied
 # node count and flavor
-openhpc_slurm_partitions: "{{ hostvars[groups['openstack'][0]]['openhpc_slurm_partitions'] }}"
+openhpc_nodegroups: "{{ hostvars[groups['openstack'][0]]['openhpc_nodegroups'] }}"

environments/.caas/inventory/group_vars/all/openondemand.yml

@@ -1,7 +1,7 @@
 ---
 openondemand_auth: basic_pam
-openondemand_jupyter_partition: "{{ openhpc_slurm_partitions[0]['name'] }}"
-openondemand_desktop_partition: "{{ openhpc_slurm_partitions[0]['name'] }}"
+openondemand_jupyter_partition: "{{ openhpc_partitions[0]['name'] }}"
+openondemand_desktop_partition: "{{ openhpc_partitions[0]['name'] }}"
 
 httpd_listen_addr_port:
   - 80

environments/.caas/inventory/group_vars/openstack.yml

@@ -16,9 +16,9 @@ terraform_project_path: "{{ playbook_dir }}/terraform"
 terraform_state: "{{ cluster_state | default('present') }}"
 cluster_ssh_user: rocky
 
-# Provision a single "standard" compute partition using the supplied
+# Provision a single "standard" compute nodegroup using the supplied
 # node count and flavor
-openhpc_slurm_partitions:
+openhpc_nodegroups:
   - name: "standard"
     count: "{{ compute_count }}"
     flavor: "{{ compute_flavor }}"

environments/common/inventory/group_vars/all/openhpc.yml

@@ -15,8 +15,21 @@ openhpc_slurmdbd_mysql_password: "{{ vault_mysql_slurm_password }}"
 openhpc_slurmdbd_mysql_username: slurm
 openhpc_slurm_control_host: "{{ groups['control'] | first }}" # avoid using hostvars for compute-init
 openhpc_slurmdbd_host: "{{ openhpc_slurm_control_host }}"
-openhpc_slurm_partitions:
-  - name: "compute"
+openhpc_rebuild_partition: # not a role var - could actually add more indirection here for things we're expecting to be modified, e.g. groups and maxtime
+  name: rebuild
+  nodegroups: "{{ cluster_compute_groups | default([]) }}"
+  default: NO
+  maxtime: 30
+  partition_params:
+    PriorityJobFactor: 65533
+    Hidden: YES
+    RootOnly: YES
+    DisableRootJobs: NO
+    PreemptMode: 'OFF'
+    OverSubscribe: EXCLUSIVE
+openhpc_nodegroups: "{{ cluster_compute_groups | map('community.general.dict_kv', 'name') }}" # create nodegroup for each compute group
+openhpc_user_partitions: "{{ openhpc_nodegroups }}" # create partition for each nodegroup (actually role default) - this is what we'd expect to be changed
+openhpc_partitions: "{{ openhpc_user_partitions + ([openhpc_rebuild_partition] if groups['rebuild'] | length > 0 else []) }}" # auto-create rebuild partition if reqd.
 openhpc_packages_default:
   # system packages
   - podman

environments/common/inventory/group_vars/all/openondemand.yml

@@ -12,8 +12,8 @@ openondemand_servername: "{{ hostvars[groups['openondemand'].0].ansible_host if
 
 openondemand_auth: basic_pam
 
-openondemand_jupyter_partition: "{{ openhpc_slurm_partitions[0]['name'] }}"
-openondemand_desktop_partition: "{{ openhpc_slurm_partitions[0]['name'] }}"
+openondemand_jupyter_partition: "{{ openhpc_partitions[0]['name'] }}"
+openondemand_desktop_partition: "{{ openhpc_partitions[0]['name'] }}"
 
 # Regex defining hosts which openondemand can proxy; the default regex is compute nodes (for apps) and grafana host,
 # e.g. if the group `compute` has hosts `compute-{0,1,2,..}` this will be '(compute-\d+)|(control)'.

environments/skeleton/{{cookiecutter.environment}}/tofu/inventory.tf

@@ -11,12 +11,3 @@ resource "local_file" "hosts" {
                           )
   filename = "../inventory/hosts.yml"
 }
-
-resource "local_file" "partitions" {
-    content  = templatefile("${path.module}/partitions.tpl",
-                            {
-                              "compute_groups": module.compute,
-                            },
-    )
-    filename = "../inventory/group_vars/all/partitions.yml" # as all/ is created by skeleton
-}

environments/skeleton/{{cookiecutter.environment}}/tofu/inventory.tpl

@@ -2,6 +2,7 @@ all:
     vars:
         openhpc_cluster_name: ${cluster_name}
         cluster_domain_suffix: ${cluster_domain_suffix}
+        cluster_compute_groups: ${jsonencode(keys(compute_groups))}
 
 control:
     hosts:

requirements.yml

@@ -4,7 +4,7 @@ roles:
     version: v25.3.2
     name: stackhpc.nfs
   - src: https://github.com/stackhpc/ansible-role-openhpc.git
-    version: v0.30.0
+    version: v1.0.0
     name: stackhpc.openhpc
   - src: https://github.com/stackhpc/ansible-node-exporter.git
     version: stackhpc