Pre/Post puppet task resources only generated for Controller role
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Fix Released
|
High
|
James Slagle |
Bug Description
The pre and post puppet tasks (ControllerPreP
These resources are used to manage the pacemaker cluster during stack-update operations, by putting the cluster into maintenance mode, and restarting needed resources at the end of the update via pcs. So, it means you can't have any role running pacemaker services that isn't specifically called Controller.
This is an issue when using the deployed-server environment and custom roles. We have to use a custom role *not* named Controller so that we can set disable_constraints on the role, but we also need to have these task resources present so that we can do the needed operations on the pacemaker cluster during stack-updates.
Changed in tripleo: | |
importance: | Undecided → High |
assignee: | nobody → James Slagle (james-slagle) |
milestone: | none → pike-1 |
tags: | added: ocata-backport-potential |
Changed in tripleo: | |
milestone: | pike-1 → ocata-rc2 |
Fix proposed to branch: master /review. openstack. org/434451
Review: https:/