First checkin of slurm-trigger role. This adds a slurm-trigger to your current...

First checkin of slurm-trigger role. This adds a slurm-trigger to your current running slurm-trigger
See README.rst for usage instructions


Former-commit-id: 052064ff

roles/slurm-trigger/README.rst

+THis role sets up trigger events on your slurm cluster.
+What you want the triggers to do is up to you, so you will probably modify the templated shell files.
+Copy the role to a local role directory?
+
+Triggers used in this role as it stands
+- primary_slurmctld_failure
+- primary_slurmctld_resumed_operation.sh
+- node down
+
+USAGE:
+- hosts: 'ManagementNodes'
+  tasks:
+   - include_vars: vars/slurm.yml
+
+- hosts: 'ManagementNodes'
+  roles:
+   - slurm_trigger  
+
+
+The role uses several variables that need to be defined:
+{{ slurm_dir }}		The directory of slurm install. Shell scripts are copied to sbin
+{{ admin_email }} 	Email address (defined in slurm.yml, or defined some other way) to send alerts to
+
+Each trigger has 2 files. One to respond to a trigger. And one to reset the trigger. The role calls the last one to start the process.
+

roles/slurm-trigger/tasks/main.yml

+---
+############################
+- name:  template primary_slurmctld_failure
+  template: dest="{{ slurm_dir }}/sbin/primary_slurmctld_failure.sh" src=primary_slurmctld_failure.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+- name:  template set primary_slurmctld_failure trigger
+  template: dest="{{ slurm_dir }}/sbin/set_primary_slurmctld_failure_trigger.sh" src=set_primary_slurmctld_failure_trigger.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+- name: Execute set_primary_slurmctld_failure)trigger  
+  command: "{{ slurm_dir }}/sbin/set_primary_slurmctld_failure_trigger.sh"
+  become: true
+  become_user: slurm
+  run_once: true
+ 
+- name:  template primary_slurmctld_resumed_operation
+  template: dest="{{ slurm_dir }}/sbin/primary_slurmctld_resumed_operation.sh" src=primary_slurmctld_resumed_operation.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+- name:  template set primary_slurmctld_resumed trigger
+  template: dest="{{ slurm_dir }}/sbin/set_primary_slurmctld_resumed_operation_trigger.sh" src=set_primary_slurmctld_resumed_operation_trigger.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+- name: Execute primary_slurmctld_resumed_operation.sh  
+  command: "{{ slurm_dir }}/sbin/set_primary_slurmctld_resumed_operation_trigger.sh"
+  become: true
+  become_user: slurm
+  run_once: true
+  
+- name:  template node_down
+  template: dest="{{ slurm_dir }}/sbin/node_down.sh" src=node_down.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+- name:  template node_down trigger command
+  template: dest="{{ slurm_dir }}/sbin/set_node_trigger.sh" src=set_node_trigger.sh.j2 mode="0755"
+  become: true
+  become_user: root
+
+
+- name: Execute set_node_trigger.sh
+  command: "{{ slurm_dir }}/sbin/set_node_trigger.sh"
+  become: true
+  become_user: slurm
+  run_once: true
+  
+ 
+ 

roles/slurm-trigger/templates/node_down.sh.j2

+#!/bin/bash
+# Notify the administrator of the failure using by e-mail
+echo "On `hostname`:`date`:`whoami`: slurm-trigger event for NODE_FAILURE: $*" | `which mail` -s "NODE FAILURE $*" {{ admin_email }}
+# Submit trigger for next primary slurmctld failure event
+TRIGGER_CMD="{{ slurm_dir }}/sbin/set_node_trigger.sh"
+
+FILE=/tmp/node_down.txt
+#COMMAND="su slurm -c $TRIGGER_CMD"
+echo "node_down.sh: `date`: `whoami`: $TRIGGER_CMD" >> $FILE
+$TRIGGER_CMD >> $FILE 2>&1

roles/slurm-trigger/templates/primary_slurmctld_failure.sh.j2

+#!/bin/bash
+# Notify the administrator of the failure using by e-mail
+echo "On `hostname`:`date`:`who`: slurm-trigger event for Primary_SLURMCTLD_FAILURE" | `which mail` -s Primary_SLURMCTLD_FAILURE {{ admin_email }} 
+# Submit trigger for next primary slurmctld failure event
+TRIGGER_CMD="{{ slurm_dir }}/sbin/set_primary_slurmctld_failure_trigger.sh"
+
+FILE=/tmp/primary_down.txt
+#COMMAND="su slurm -c $TRIGGER_CMD"
+echo "primary_slurmctld_failure.sh:`date`:`whoami`: $TRIGGER_CMD" >> $FILE
+$TRIGGER_CMD >> $FILE 2>&1

roles/slurm-trigger/templates/primary_slurmctld_resumed_operation.sh.j2

+#!/bin/bash
+# Notify the administrator of the failure using by e-mail
+echo "On `hostname`:`date`:`whoami`: slurm-trigger event for Primary_SLURMCTLD_RESUMED" | `which mail` -s Primary_SLURMCTLD_RESUMED {{ admin_email }} 
+# Submit trigger for next primary slurmctld failure event
+
+FILE=/tmp/primary_up.txt
+#COMMAND="su slurm -c {{ slurm_dir }}/sbin/set_primary_slurmctld_resumed_operation_trigger.sh"
+COMMAND="{{ slurm_dir }}/sbin/set_primary_slurmctld_resumed_operation_trigger.sh"
+echo "primary_slurmctld_resumed_operation.sh.sh:`date`:`whoami`: $COMMAND" >> $FILE
+$COMMAND >> $FILE 2>&1

roles/slurm-trigger/templates/set_node_trigger.sh.j2

+#!/bin/bash
+TRIGGER_CMD="{{ slurm_dir }}/bin/strigger --set --down  --program={{ slurm_dir }}/sbin/node_down.sh"
+echo "set_node_trigger.sh: `date`: $TRIGGER_CMD" 
+$TRIGGER_CMD 

roles/slurm-trigger/templates/set_primary_slurmctld_failure_trigger.sh.j2

+#!/bin/bash
+TRIGGER_CMD="{{ slurm_dir }}/bin/strigger --set --primary_slurmctld_failure  --program={{ slurm_dir }}/sbin/primary_slurmctld_failure.sh"
+echo "set_primary_slurmctld_failure_trigger.sh: `date`: $TRIGGER_CMD"
+$TRIGGER_CMD

roles/slurm-trigger/templates/set_primary_slurmctld_resumed_operation_trigger.sh.j2

+#!/bin/bash
+TRIGGER_CMD="{{ slurm_dir }}/bin/strigger --set --primary_slurmctld_resumed_operation  --program={{ slurm_dir }}/sbin/primary_slurmctld_resumed_operation.sh"
+echo "set_primary_slurmctld_resumed_operation_trigger.sh: `date`: $TRIGGER_CMD"
+$TRIGGER_CMD