Failed to start SLAPD and B&R operation failed in ansible

Bug #2017927 reported by Kirti Singh
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
StarlingX
Fix Released
Medium
Kirti Singh

Bug Description

Ldap conf is pointing to itself as the provider in sync-replication directive hence the restore is failing.

Getting the below error while executing ansible playbook for restore.
stdout: 'Starting SLAPD: Failed to start SLAPD.

affects: update-manager (Ubuntu) → openldap (Ubuntu)
Kirti Singh (kirtis)
affects: openldap (Ubuntu) → starlingx
Kirti Singh (kirtis)
Changed in starlingx:
assignee: nobody → Kirti Singh (kirtis)
Changed in starlingx:
status: New → In Progress
Revision history for this message
John Kung (john-kung) wrote :

Appending to Bug Description (from review) https://review.opendev.org/c/starlingx/ansible-playbooks/+/882116

"In a Duplex system when backup is taken from
controller-1 and performed restore on controller-0, the
restore fails due to SLAPD failure.

SLAPD fails to start because, the backup file
/etc/ldap/schema/cn=config/olcDatabase={1}mdb.ldif defines
the value, provider=ldaps://controller-0 in the
replication-sync(olcSyncrepl) parameter(directive)."

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to ansible-playbooks (master)

Reviewed: https://review.opendev.org/c/starlingx/ansible-playbooks/+/882116
Committed: https://opendev.org/starlingx/ansible-playbooks/commit/8c7ff12aff7073be45c88d6fe312b54d4d854f05
Submitter: "Zuul (22348)"
Branch: master

commit 8c7ff12aff7073be45c88d6fe312b54d4d854f05
Author: ksingh <email address hidden>
Date: Wed May 3 02:57:08 2023 -0400

    Fix ldap restore failure when backup taken from controller-1

    The issue is, in a Duplex system when backup is taken from
    controller-1, the restore fails due to SLAPD failure.

    SLAPD fails to start because, the backup file
    /etc/ldap/schema/cn=config/olcDatabase={1}mdb.ldif defines
    the value, provider=ldaps://controller-0 in the
    replication-sync(olcSyncrepl) parameter(directive).

    When we do restore from controller-0 the provider ID
    controller-1 is expected.

    NOTE: Restore has to be from controller-0 as for the restore
    operation, controller-0 is implicit. Since restore would be
    used in disaster recovery/install scenario on duplex, restore
    is expected on controller-0.

    As a solution, conversion of provider value in sync-replication
    directive(olcsyncrepl) is made to point to controller-1.

    In AIO-SX, controller-0 provider is set to controller-1. This is
    the default value for AIO-SX. The changes mentioned in this review
    does not change the value for AIO-SX Backup and Restore.

    TEST PLAN
    PASS: AIO-SX: Backup and Restore passed
    PASS: AIO-DX: Backup and Restore passed

    Closes-Bug: 2017927

    Change-Id: Ib6c7655e8a8e0fbe8d1426918839054de44ebe64
    Signed-off-by: ksingh <email address hidden>

Changed in starlingx:
status: In Progress → Fix Released
Ghada Khalil (gkhalil)
tags: added: stx.9.0 stx.update
Changed in starlingx:
importance: Undecided → Medium
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.