Reviewed:  
https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/1005240
Committed: 
https://opendev.org/openstack/openstack-ansible-os_neutron/commit/0b06045d1032d4a4e89bd524fc6c8517f5eeee49
Submitter: "Zuul (22348)"
Branch:    master

commit 0b06045d1032d4a4e89bd524fc6c8517f5eeee49
Author: Dmitriy Rabotyagov <[email protected]>
Date:   Fri Sep 11 15:52:35 2026 +0200

    Statically provision ovn-central parameters
    
    At the moment there is an ambiguous behavior when OVN Leader is being
    re-deployed, which can be part of the OS upgrade.
    According to the logic, when we shut down the leader, it is re-elected
    by the raft cluster, so `_check_cluster_db` exists with non-zero code,
    but leader_node has non-zero length.
    
    This results in ovn-central provisioned with remote-addr, which means
    that database upgrade does not happen.
    At worst case scenario, this may end up in nodes forming an
    independent cluster of their own, not joining anywhere and basically
    nuking existing networking.
    
    Thus, we add another logical condition to handle node re-joining and
    bootstrapping a new cluster as two differnet actions, where ovn-central
    config file gets temporary remote-addr to join the cluster, which then
    re-written if this supposed to be the leader node.
    
    Closes-Bug: #2059721
    Change-Id: I5cab1fe24e8c88fe8961e3cf2bca45ad14bc1918
    Signed-off-by: Dmitriy Rabotyagov <[email protected]>


** Changed in: openstack-ansible
       Status: In Progress => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2059721

Title:
  ovn db upgrade

To manage notifications about this bug go to:
https://bugs.launchpad.net/openstack-ansible/+bug/2059721/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to