Merge pull request #201 from stackhpc/feat/template-mpi-conf

Add support for mpi.conf templating

Author: bertiethorpe

README.md

@@ -126,6 +126,13 @@ used to supplement or override the template defaults. Templated parameters can
 also be removed by setting the value to the literal string `'omit'` - note
 that this is *not the same* as the Ansible `omit` [special variable](https://docs.ansible.com/ansible/latest/reference_appendices/special_variables.html#term-omit).
 
+`openhpc_mpi_config`: Optional. Mapping of additional parameters and values for
+[mpi.conf](https://slurm.schedmd.com/mpi.conf.html). Keys are mpi.conf
+parameter names and values are lists or strings as appropriate. This can be
+used to supplement or override the template defaults. Templated parameters can
+also be removed by setting the value to the literal string `'omit'` - note
+that this is *not the same* as the Ansible `omit` [special variable](https://docs.ansible.com/ansible/latest/reference_appendices/special_variables.html#term-omit).
+
 `openhpc_ram_multiplier`: Optional, default `0.95`. Multiplier used in the calculation: `total_memory * openhpc_ram_multiplier` when setting `RealMemory` for the partition in slurm.conf. Can be overriden on a per partition basis using `openhpc_slurm_partitions.ram_multiplier`. Has no effect if `openhpc_slurm_partitions.ram_mb` is set.
 
 `openhpc_state_save_location`: Optional. Absolute path for Slurm controller state (`slurm.conf` parameter [StateSaveLocation](https://slurm.schedmd.com/slurm.conf.html#OPT_StateSaveLocation))

defaults/main.yml

@@ -51,6 +51,9 @@ openhpc_cgroup_config: {}
 openhpc_gres_template: gres.conf.j2
 openhpc_cgroup_template: cgroup.conf.j2
 
+openhpc_mpi_template: mpi.conf.j2
+openhpc_mpi_config: {}
+
 openhpc_state_save_location: /var/spool/slurm
 openhpc_slurmd_spool_dir: /var/spool/slurm
 openhpc_slurm_conf_path: /etc/slurm/slurm.conf

tasks/runtime.yml

@@ -103,6 +103,20 @@
   register: ohpc_cgroup_conf
   # NB uses restart rather than reload as this is needed in some cases
 
+- name: Template mpi.conf
+  template:
+    src: "{{ openhpc_mpi_template }}"
+    dest: "{{ openhpc_slurm_conf_path | dirname }}/mpi.conf"
+    owner: root
+    group: root
+    mode: "0644"
+  when:
+    - openhpc_enable.control | default(false)
+    - openhpc_mpi_config | length > 0
+  notify:
+    - Restart slurmctld service
+  register: ohpc_mpi_conf
+
 # Workaround for https://bugs.rockylinux.org/view.php?id=10165
 - name: Fix permissions on /etc for Munge service
   ansible.builtin.file:
@@ -134,6 +148,7 @@
   when:
     - openhpc_slurm_control_host in ansible_play_hosts
     - hostvars[openhpc_slurm_control_host].ohpc_slurm_conf.changed or
+      hostvars[openhpc_slurm_control_host].ohpc_mpi_conf.changed or
       hostvars[openhpc_slurm_control_host].ohpc_cgroup_conf.changed or
       hostvars[openhpc_slurm_control_host].ohpc_gres_conf.changed # noqa no-handler
   notify:

templates/mpi.conf.j2

@@ -0,0 +1,6 @@
+# {{ ansible_managed }}
+{% for k, v in openhpc_mpi_config.items %}
+{%  if v != "omit" %}{# allow removing items using setting key: omit #}
+{{ k }}={{ v | join(',') if (v is sequence and v is not string) else v }}
+{%  endif %}
+{% endfor %}