Try to do nagios downtimes better for fmn upgrades.

This commit is contained in:
Ralph Bean 2015-03-19 15:07:32 +00:00
parent 857adabe4f
commit f26d80d3b1

View file

@ -33,7 +33,7 @@
pre_tasks: pre_tasks:
- name: tell nagios to shush w.r.t. the frontend - name: tell nagios to shush w.r.t. the frontend
nagios: action=downtime minutes=60 service=host host={{ inventory_hostname }} nagios: action=downtime minutes=15 service=host host={{ inventory_hostname }}
delegate_to: noc01.phx2.fedoraproject.org delegate_to: noc01.phx2.fedoraproject.org
ignore_errors: true ignore_errors: true
@ -55,7 +55,7 @@
pre_tasks: pre_tasks:
- name: tell nagios to shush w.r.t. the backend - name: tell nagios to shush w.r.t. the backend
nagios: action=downtime minutes=60 service=host host={{ inventory_hostname }} nagios: action=downtime minutes=15 service=host host={{ inventory_hostname }}
delegate_to: noc01.phx2.fedoraproject.org delegate_to: noc01.phx2.fedoraproject.org
ignore_errors: true ignore_errors: true
@ -76,11 +76,13 @@
- name: And... start the backend again - name: And... start the backend again
service: name="fedmsg-hub" state=started service: name="fedmsg-hub" state=started
post_tasks: # Don't bother unshushing the backend here. it takes a few minutes to start
- name: tell nagios to unshush w.r.t. the backend # up anyways, so just let the downtime expire.
nagios: action=unsilence service=host host={{ inventory_hostname }} #post_tasks:
delegate_to: noc01.phx2.fedoraproject.org #- name: tell nagios to unshush w.r.t. the backend
ignore_errors: true # nagios: action=unsilence service=host host={{ inventory_hostname }}
# delegate_to: noc01.phx2.fedoraproject.org
# ignore_errors: true
- name: restart the frontend - name: restart the frontend
hosts: notifs-web;notifs-web-stg hosts: notifs-web;notifs-web-stg