[os_keystone] Error when installing keystone with LDAP

Bug #1800425 reported by Christian Zunker
14
This bug affects 3 people
Affects Status Importance Assigned to Milestone
OpenStack-Ansible
Fix Released
Undecided
Unassigned

Bug Description

osa version: 17.1.1 on Ubuntu

We are trying to upgrade pike to queens with osa 17.1.1.
We have a three controller setup and keystone installation works on the first container, but fails on the remaining two during the task "Create Keystone LDAP domains":
SSL exception connecting to https://172.29.236.254:35357/v3/auth/tokens: HTTPSConnectionPool(host='172.29.236.254', port=35357): Max retries exceeded with url: /v3/auth/tokens (Caused by SSLError(SSLEOFError(8, u'EOF occurred in violation of protocol

The problem is the HAProxy backend. Alle keystone backends are in status MAINT. With the first container, only one container gets disabled in the backend, but not enabled again, after the keystone installation is finished for this container.

These are the tasks:
TASK [Set haproxy service state] **************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
Monday 29 October 2018 06:10:40 +0000 (0:00:00.167) 0:02:27.497 ********
 [WARNING]: The loop variable 'item' is already in use. You should set the `loop_var` value in the `loop_control` option for the task to something else to avoid variable collisions and unexpected behavior.

ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.23] => (item=ctr003)
ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.14] => (item=ctr002)
ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.16] => (item=ctr001)

TASK [Set haproxy service state] **************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
Monday 29 October 2018 06:10:44 +0000 (0:00:03.771) 0:02:31.268 ********
 [WARNING]: The loop variable 'item' is already in use. You should set the `loop_var` value in the `loop_control` option for the task to something else to avoid variable collisions and unexpected behavior.

ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.23] => (item=ctr003)
ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.14] => (item=ctr002)
ok: [ctr003_keystone_container-2bdcda67 -> 172.20.20.16] => (item=ctr001)

My guess is, this only happens when tasks are executed which need access to keystone while keystone is installed.
The only other hint I found is about cinder:
http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/%23openstack-ansible.2018-03-23.log.html#t2018-03-23T01:20:42

The above problem might be gone, when this is merged:
https://review.openstack.org/#/c/504795/

Revision history for this message
Christian Zunker (christian-zunker) wrote :
Revision history for this message
Mohammed Naser (mnaser) wrote :
Changed in openstack-ansible:
status: New → Fix Committed
Revision history for this message
Jesse Pretorius (jesse-pretorius) wrote :

Actually, it was released in 17.1.2

Changed in openstack-ansible:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.