Uploaded image for project: 'RHEL'
  1. RHEL
  2. RHEL-83274

Replication online reinitialization of a large database gets stalled.

Linking RHIVOS CVEs to...Migration: Automation ...SWIFT: POC ConversionSync from "Extern...XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Blocker Blocker
    • rhel-10.2
    • rhel-9.4, rhel-9.5
    • 389-ds-base
    • None
    • None
    • Moderate
    • ZStream
    • rhel-idm-ds
    • ssg_idm
    • None
    • None
    • Bug Fix
    • Hide
      Cause: parentid index is not using the proper matching rule
      Consequence: Replication online init is failing as the parentid index can not be used to build the list of entries.
      Fix: The parentid index needs to be created with integer matching rule. If it already exist with improper MR, healthcheck tool reports it.
      Result: Once parentid is indexed properly, the supplier can complete an online total init.
      Show
      Cause: parentid index is not using the proper matching rule Consequence: Replication online init is failing as the parentid index can not be used to build the list of entries. Fix: The parentid index needs to be created with integer matching rule. If it already exist with improper MR, healthcheck tool reports it. Result: Once parentid is indexed properly, the supplier can complete an online total init.
    • Proposed
    • None

      What were you trying to do that didn't work?

      Customers are trying to reinitialize replicas from a supplier that has a large database ( typically over the million of entries ).
      This is failing after several minutes with the message "import <BACKEND_NAME>: Thread monitoring returned: -23".
      The import rate seems too high to be correct and the process seems to not progress after trying several passes.

      What is the impact of this issue to you?

      This forces customers to stop their suppliers to use the offline method.

      Please provide the package NVR for which the bug is seen:

      cat etc/redhat-release
Red Hat Enterprise Linux release 9.4 (Plow)grep 389-ds installed-rpms
389-ds-base-2.4.5-9.el9_4.x86_64                            Fri Oct 11 07:40:02 2024
389-ds-base-libs-2.4.5-9.el9_4.x86_64                       Fri Oct 11 07:40:01 2024
cockpit-389-ds-2.2.9-2.el9.noarch                           Fri Oct 11 07:40:04 2024

      How reproducible is this bug?:

      Quite consistently with large databases ( over the million of entries ).

      Steps to reproduce

      1. Create 2 RHDS instances and configure replication
      2. Add millions of entries to the supplier
      3. Try to reinitialize the consumer with the online method ( CLI or Console )

      Expected results

      Working online reinitialization.

      Actual results

      Failing to leverage the online reinitialization method.

              tbordaz@redhat.com Thierry Bordaz
              rhn-support-tmihinto Têko Mihinto
              IdM DS Dev IdM DS Dev
              IdM DS QE IdM DS QE
              Evgenia Martyniuk Evgenia Martyniuk
              Votes:
              1 Vote for this issue
              Watchers:
              13 Start watching this issue

                Created:
                Updated: