Handlers are not delegated properly for periodic nova db archiving
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack-Ansible |
Fix Released
|
Critical
|
Dmitriy Rabotyagov |
Bug Description
Problem description:
In [1] we implemented a feature that allows users to periodically trim old database records in nova db.
Unfortunately, when this feature is enabled, handler from systemd_service role is not delegated properly to the first nova_conductor host and it's executed on a compute node instead.
It's not expected to have this timer on a compute node, so the playbook fails with an error:
```
RUNNING HANDLER [systemd_service : Restart service {{ services_
failed: [aio1] (item=) => {"ansible_
```
Steps to reproduce:
1. Define `nova_archive_
2. Execute `openstack-ansible /opt/openstack-
Proposed solution:
So far I haven't been able to find any solution.
description: | updated |
summary: |
- Handlers not delegated properly for periodic nova db archiving + Handlers are not delegated properly for periodic nova db archiving |
ansible- role-systemd_ service/ tasks/main. yml deleted. timer configuration was not generated when restarting the service。
The trigger processor is used in line 88. The nova-archive-