Back to bug 2225205

Who When What Removed Added
Luca Miccini 2023-07-24 14:44:07 UTC Keywords Triaged
Priority unspecified high
Assignee rhos-maint lmiccini
Status NEW ON_DEV
Red Hat One Jira (issues.redhat.com) 2023-07-24 14:45:33 UTC Link ID Red Hat Issue Tracker OSP-26862
Luca Miccini 2023-07-24 14:47:10 UTC Link ID OpenStack gerrit 889306
Luca Miccini 2023-07-31 16:35:37 UTC Status ON_DEV POST
Luca Miccini 2023-08-01 07:27:17 UTC Fixed In Version openstack-tripleo-heat-templates-14.3.1-17.1.20230801070949.d134056.el9osttrunk
Status POST MODIFIED
Luca Miccini 2023-08-01 07:40:24 UTC Target Milestone --- z1
RHEL Program Management 2023-08-01 07:40:35 UTC Target Release --- 17.1
Paul Grist 2023-08-01 19:37:42 UTC CC pgrist
Ian Frangs 2023-08-03 15:46:23 UTC Flags needinfo?(lmiccini)
Luca Miccini 2023-08-03 16:08:42 UTC Flags needinfo?(lmiccini)
Doc Type If docs needed, set a value Known Issue
Doc Text Cause:

Outdated upgrade orchestration logic will override the existing pacemaker authkey during the Fast Forward Upgrade, preventing pacemaker from connecting to pacemaker_remote running on compute nodes when Instance HA is enabled.

Consequence:

Fast Forward Upgrade will fail and pacemaker_remote running on compute nodes will be unreachable from the central cluster.

Workaround (if any):

Reach out to Red Hat support to receive instructions on how to perform FFU if Instance HA is configured.

Result:

Fast Forward Upgrade will fail and compute nodes could be potentially end up in a fence loop.
Joanne O'Flynn 2023-08-04 11:02:34 UTC Doc Text Cause:

Outdated upgrade orchestration logic will override the existing pacemaker authkey during the Fast Forward Upgrade, preventing pacemaker from connecting to pacemaker_remote running on compute nodes when Instance HA is enabled.

Consequence:

Fast Forward Upgrade will fail and pacemaker_remote running on compute nodes will be unreachable from the central cluster.

Workaround (if any):

Reach out to Red Hat support to receive instructions on how to perform FFU if Instance HA is configured.

Result:

Fast Forward Upgrade will fail and compute nodes could be potentially end up in a fence loop.
Outdated upgrade orchestration logic overrides the existing pacemaker authkey during the Fast Forward Upgrade (FFU) procedure, preventing Pacemaker from connecting to `pacemaker_remote` running on Compute nodes when Instance HA is enabled. As a result, the upgrade fails and `pacemaker_remote` running on Compute nodes is unreachable from the central cluster. Contact Red Hat support to receive instructions on how to perform FFU if Instance HA is configured.
CC joflynn

Back to bug 2225205