Role - edpm_reboot

Usage

This Ansible role allows to do the following tasks:

  • Check if EDPM compute reboot is required. It takes care of checking EDPM hosts to see if other roles set flag to indicate need of reboot and if needs-restarting yum plugin reported that reboot is required.

  • Reboot EDPM computes. During deployment reboot is triggered automatically if required. During post-deployment reconfiguration or adoption process reboot is not started. User has to plan reboot maintenance window and set edpm_reboot_strategy flag to force.

Here is an example of a playbook to force start reboot:

- name: Force start reboot of nodes
  block:
    - name: "Force start reboot of nodes"
      include_role:
        name: edpm_reboot
       vars:
         edpm_reboot_strategy: force

osp.edpm.edpm_reboot role – The main entry point for the edpm_reboot role.

Entry point main – The main entry point for the edpm_reboot role.

Synopsis
Parameters

Parameter

Comments

edpm_reboot_edpm_node_config_dir

path

Path for storing configuration of edpm nodes. This should be synced with edpm-ansible roles.

Default: "/var/lib/openstack/config"

edpm_reboot_old_tripleo_node_config_dir

path

Path to check for tripleo pre-adopted nodes

Default: "/var/lib/config-data/puppet-generated"

edpm_reboot_post_reboot_delay

integer

Reboot delay in seconds

Default: 60

edpm_reboot_strategy

string

Default strategy is auto. In auto mode reboot for deployed nodes is by default defered as it can impact running vms or openstack services. Only on initial run when nova related files are not yet created reboot is not defered. When edpm_reboot_force_reboot is set to force, reboot is allowed and will be perfomed always.

Default: "auto"

edpm_reboot_timeout_reboot

integer

Reboot timeout in seconds

Default: 3600