Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 2186152

Summary: Suboptimal approach is used when choosing IPA_HOST from IdMServer list
Product: Red Hat OpenStack Reporter: Alex Stupnikov <astupnik>
Component: openstack-tripleo-heat-templatesAssignee: Dave Wilde <dwilde>
Status: CLOSED DUPLICATE QA Contact: Joe H. Rahme <jhakimra>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 16.2 (Train)CC: chopark, dwilde, ggrasza, mburns, tkajinam
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: All   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2023-05-08 10:47:12 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Alex Stupnikov 2023-04-12 08:49:11 UTC
Description of problem:

Customer reported a problem with deployment framework: deployment fails when first IdM server from IdmServer list is unreachable, but deployment doesn't fail when other servers are unreachable:

2023-04-11 12:55:31,919 p=99490 u=mistral n=ansible | 2023-04-11 12:55:31.919475 | 525400ab-8551-faf2-6150-000000002e89 |      FATAL | add sub_host | undercloud | error={"changed": false, "msg": "host_find: Request failed: <urlopen error [Errno 113] No route to host>"}

It looks like this problem is caused by logic of IPA_HOST selection in deployment/ipa/ipaservices-baremetal-ansible.yaml: THT always takes first IdM server from the list:

            if:
              - idm_server_provided
              - IPA_HOST: {get_param: [IdMServer, 0]}

This doesn't look like a major problem, there is a workaround (to tune order of IdMServer entries). But at the same time we expect some form of HA when multiple IdM servers are configured and one of them is down.

Version-Release number of selected component (if applicable):
RHOSP 16.2

How reproducible:
Deployment fails when communication between director and first IdM server from IdMServer list doesn't work

Actual results:
Deployment fails

Expected results:
THT failovers to available IdM servers

Comment 2 Grzegorz Grasza 2023-05-08 10:47:12 UTC
The issue stated here is because we implemented a partial fix, for supplying a list of IDM servers, the followup work is being tracked in another bugzilla.

I'm closing this as a duplicate, see my comment here on the other bug: https://bugzilla.redhat.com/show_bug.cgi?id=2034765#c3

*** This bug has been marked as a duplicate of bug 2034765 ***