token flush cronjob not being enabled on leader unit
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Keystone Charm |
Fix Released
|
Medium
|
James Page |
Bug Description
I have seen in a few xenial-queens clouds that the token flush cronjob is completely disabled:
jujumanage@MAAS:~$ juju run --application keystone -- 'is-leader; cat /etc/cron.
- Stdout: |
False
# Purge expired tokens from the keystone database hourly, per OpenStack installation guide.
# Current unit is not the leader unit. Token flush will be managed by the leader unit.
UnitId: keystone/0
- Stdout: |
False
# Purge expired tokens from the keystone database hourly, per OpenStack installation guide.
# Current unit is not the leader unit. Token flush will be managed by the leader unit.
UnitId: keystone/1
- Stdout: |
True
# Purge expired tokens from the keystone database hourly, per OpenStack installation guide.
# Current unit is not the leader unit. Token flush will be managed by the leader unit.
UnitId: keystone/2
The above happens with charm versions 289, 290, and 291.
The cronjob does however work fine in a xenial-ocata cloud using the 18.05 series, so the bug has probably been introduced somewhere between 18.05 and 18.11
tags: | added: canonical-bootstack |
Changed in charm-keystone: | |
status: | Fix Committed → Fix Released |
It happens on xenial-pike with keystone v285 as well.