Merge pull request dell#178 from VishnupriyaKrish/devel

Issue dell#172: Slurm role changes and testing framework for Slurm

omnia.yml

@@ -57,12 +57,6 @@
     - k8s_common
   tags: kubernetes
 
-- name: Apply GPU node config
-  hosts: gpus
-  gather_facts: false
-  roles:
-    - compute_gpu
-
 - name: Apply K8s manager config
   hosts: manager
   gather_facts: true

roles/slurm_common/files/slurm.conf

@@ -32,6 +32,7 @@ ProctrackType=proctrack/pgid
 #FirstJobId=
 ReturnToService=2
 #MaxJobCount=
+MailProg=/usr/bin/mail
 #PlugStackConfig=
 #PropagatePrioProcess=
 #PropagateResourceLimits=
@@ -87,11 +88,11 @@ AccountingStorageType=accounting_storage/slurmdbd
 #AccountingStorageLoc=
 #AccountingStoragePass=
 #AccountingStorageUser=
-#
+AccountingStoragePort=
 # COMPUTE NODES
 #NodeName=linux[1-32] Procs=1 State=UNKNOWN
 #NodeName=DEFAULT Sockets=2 CoresPerSocket=20 State=UNKNOWN
 NodeName= Sockets= CoresPerSocket=
 #NodeName=compute[002-005] CoresPerSocket=20
 PartitionName=normal Nodes=ALL Default=YES MaxTime=INFINITE State=UP
-#PartitionName=debug Nodes=ALL Default=YES MaxTime=INFINITE State=UP
+#PartitionName=debug Nodes=ALL Default=YES MaxTime=INFINITE State=UP

roles/slurm_common/tasks/main.yml

@@ -13,38 +13,40 @@
 #  limitations under the License.
 ---
 
-- name: Install epel repository
-  package:
-    name: epel-release
-    state: present
-  tags: install
+- name: Get hostname
+  command: hostname -s
+  register: host_name
+  changed_when: true
 
-- name: Munge installation
-  package:
-    name: munge-devel
+- name: Add host name in file
+  replace:
+    dest: "{{ hostname_dest }}"
+    regexp: localhost.localdomain
+    replace: "{{ host_name.stdout }}"
+    backup: yes
+    mode: "{{ common_mode }}"
+
+- name: Add host name in hosts file
+  lineinfile:
+    dest: "{{ hosts_dest }}"
+    line: "{{ hostvars[inventory_hostname]['ansible_default_ipv4']['address'] }} {{ host_name.stdout }}"
     state: present
-  tags: install
+    create: yes
+    mode: "{{ common_mode }}"
 
 - name: Install packages for slurm
   package:
-    name: "{{ common_packages }}"
+    name: "{{ item }}"
     state: present
+  with_items:
+    - "{{ common_packages }}"
   tags: install
 
-- name: pip upgrade pip
-  pip:
-    name: pip
-    executable: pip3
-    extra_args: --upgrade
-    state: latest
-  tags: install
-
-- name: create munge key
+- name: Create munge key
   command: "{{ munge_cmd }}"
   changed_when: true
-  tags: install
 
-- name: copy munge key
+- name: Copy munge key
   copy:
     src: munge.key
     dest: "{{ munge_dest }}"
@@ -53,75 +55,72 @@
     mode: "{{ munge_mode }}"
   tags: install
 
-- name: slurm configuration - slurm.conf
+- name: Slurm configuration - slurm.conf
   copy:
     src: slurm.conf
     dest: "{{ slurm_dest }}"
     mode: "{{ slurm_mode }}"
   tags: install
 
-- name: add cluster name
+- name: Add cluster name
   lineinfile:
     path: "{{ slurm_confpth }}"
-    regexp: "clustername="
-    line: "clustername={{ cluster_name }}"
-  tags: install
+    regexp: "ClusterName="
+    line: "ClusterName={{ cluster_name }}"
 
-- name: add slurm user name
+- name: Add slurm user name
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmUser="
     line: "SlurmUser={{ slurm_user }}"
-  tags: install
 
 - name: Add slurmctld port no
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmctldPort="
     line: "SlurmctldPort={{ slurmctld_port }}"
-  tags: install
 
 - name: Add slurmd port no
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmdPort="
     line: "SlurmdPort={{ slurmd_port }}"
-  tags: install
 
 - name: Add spool path
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmdSpoolDir="
     line: "SlurmdSpoolDir={{ spool_pth }}"
-  tags: install
 
 - name: Add slurmctld pid file path
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmctldPidFile="
     line: "SlurmctldPidFile={{ slurmctld_pid }}"
-  tags: install
 
 - name: Add slurmd pid file path
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmdPidFile="
     line: "SlurmdPidFile={{ slurmd_pid }}"
-  tags: install
 
 - name: Add slurmctld log file path
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmctldLogFile="
     line: "SlurmctldLogFile={{ slurmctld_log }}"
-  tags: install
 
 - name: Add slurmd log file path
   lineinfile:
     path: "{{ slurm_confpth }}"
     regexp: "SlurmdLogFile="
     line: "SlurmdLogFile={{ slurmd_log }}"
-  tags: install
+
+- name: Add accounting storage port no
+  lineinfile:
+    path: "{{ slurm_confpth }}"
+    regexp: "AccountingStoragePort="
+    line: "AccountingStoragePort={{ acct_port }}"
 
 - name: Create slurm group
   group:
@@ -155,6 +154,15 @@
     state: directory
     mode: "{{ gen_mode }}"
     recurse: yes
+
+- name: Create slurm pid directory
+  file:
+    path: "{{ slurm_pidpth }}"
+    state: directory
+    owner: slurm
+    group: slurm
+    mode: "{{ gen_mode }}"
+    recurse: yes
   tags: install
 
 - name: Give slurm user permission to slurmctld
@@ -164,7 +172,6 @@
     group: slurm
     mode: "{{ gen_mode }}"
     state: touch
-  tags: install
 
 - name: Give slurm user permission to slurmd
   file:
@@ -173,11 +180,11 @@
     group: slurm
     mode: "{{ gen_mode }}"
     state: touch
-  tags: install
 
 - name: Start munge service
-  service:
+  systemd:
     name: munge
     state: restarted
     enabled: yes
   tags: install
+  ignore_errors: yes

roles/slurm_common/vars/main.yml

@@ -13,31 +13,37 @@
 #  limitations under the License.
 ---
 
-epel_url: https://dl.fedoraproject.org/pub/epel/epel-release-latest-8.noarch.rpm
+epel_url: https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
 
 common_packages:
    - munge
    - munge-libs
+   - munge-devel
    - mariadb-server
    - mariadb-devel
-   - python3
-   - python-pip
+   - man2html
+   - MySQL-python
 
+hostname_dest: "/etc/hostname"
+hosts_dest: "/etc/hosts"
 munge_dest: "/etc/munge/"
 munge_cmd: "/usr/sbin/create-munge-key -f"
 munge_mode: "0400"
 slurm_mode: "0644"
+common_mode: "0777"
 slurm_dest: "/etc/slurm/"
 slurm_confpth: "/etc/slurm/slurm.conf"
 slurm_user: "slurm"
 slurmctld_port: "6817"
 slurmd_port: "6818"
+acct_port: "6819"
 slurm_uid: "6001"
 slurm_logpth: "/var/log/slurm/"
+slurm_pidpth: "/var/run/slurm/"
 gen_mode: "0755"
 spool_pth: "/var/spool/slurm/"
 slurmctld_pid: "/var/run/slurmctld.pid"
 slurmd_pid: "/var/run/slurmd.pid"
 cluster_name : "manager,compute"
 slurmctld_log: "/var/log/slurm/slurmctld.log"
-slurmd_log: "/var/log/slurm/slurmd.log"
+slurmd_log: "/var/log/slurm/slurmd.log"

roles/slurm_manager/files/slurmdbd.conf

@@ -18,7 +18,7 @@ AuthType=auth/munge
 # slurmDBD info
 DbdAddr=
 DbdHost=
-#DbdPort=7031
+#DbdPort=6019
 SlurmUser=
 #MessageTimeout=300
 DebugLevel=verbose
@@ -33,6 +33,6 @@ PidFile=
 StorageType=accounting_storage/mysql
 #StorageHost=
 #StoragePort=
-#StoragePass=
-#StorageUser=
+StoragePass=
+StorageUser=
 #StorageLoc=