How to roll back to a previous CF version in Kubernetes or OpenShift
Applies to
HCL Digital Experience v9.5 and higher
Introduction
If the upgrade of the DX core pod fails and enters a crash loop, you can revert to the previous cumulative fix (CF) version. This article describes how to roll back to a previous CF version.
Instructions
To roll back to a previous CF version when a previous version profile is present in the Persistent Volume (PV), perform the following steps:
Verifying the profile directory
-
Access the core PV.
-
Check if the
prof_95/<oldCF>directory exists.
If the directory exists, proceed to the deployment rollback steps.
Note
If the previous version profile does not exist, open a case with HCL Support to investigate the issue.
Rolling back the deployment
-
Set the replicas of the
dx-corepod to0, and then perform a Helm upgrade. -
Delete the new CF directory from the PV.
-
Set the images and tags to the old CF version, and then perform a Helm upgrade.
-
Set the replicas of the
dx-corepod to1, and then perform a Helm upgrade.