Backup and Recovery Procedures
This section shows the deployment architecture and provides the instructions to create and manage backup and recovery of HCL Digital Experience components in containerized DX 9.5 environments.
Instructions to back up the Digital Experience 9.5 Container components
The following sections describe how the administrators can create and manage backups and recovery of DX 9.5 Container components such as wp_profile
, persistence layer - database, and the media in Digital Asset Management.
-
wp-profile backup
-
Backup the file systems in the Digital Experience 9.5 container profile (Persistent volume claim
wp_profile
). Refer to the Backup and restore topic and component backup guidance for more information.Note
The HCL Digital Asset Management (DAM) component uploads folder and the DAM persistent mount for the primary instance dx-deployment-persistence-0 in the statefulset.
To create a backup of the profile Persistent volume claim
wp_profile
, it is recommended that:- The DX has only one instance.
-
The DX 9.5 container instance is stopped using the HCL Portal and HCL Web Content Manager command as follows:
kubectl exec --stdin --tty <DX_POD_NAME> -n <NAMESPACE> -- /bin/bash cd /opt/HCL/wp_profile/bin/ /stopServer.sh WebSphere_Portal -username <USERNAME> -password <PASSWORD>
-
The entire /opt/HCL/wp_profile directory is backed up.
In the command line interface, run the following command to back up the Digital Experience 9.5 Persistent volume claim wp_profile:
-
Before running the tar command, ensure that the backup file system that you are using has ~50% free profile.
cd /opt/HCL/wp_profile tar -cvpzf backup.tar.gz --exclude=/backup.tar.gz --one-file-system /opt/HCL/wp_profile/*
After the tar backup command is completed, it is recommended that a copy of the backup.tar.gz file is created and placed to alternate long term storage.
-
Recover from wp_profile pervasive volume backup
You can extract the backup of the wp_profile volume to recover any files you need to restore.
If the version of the backup matches the current fixpack level, you can use the extracted files to populate the original pervasive volume.
The procedure to do this depends on how the backup was created.
Any changes that occurred after the backup was created will not be recovered.
The portal database must be restored to the backup that was created when the backup of wp_profile was created.
-
-
Persistence layer - database backup
Run the following command to back up the container components that are managed through the DX Persistence layer:
pg_dump name_of_database > name_of_backup_file
To back up the system components on a remote system:
pg_dump -U user_name -h remote_host -p remote_port name_of_database > name_of_backup_file
After the backup command is completed, it is recommended that a copy of the resulting file is created and placed to an alternate long term storage.
See Backup and restore a DAM image for more information.
-
Digital Asset Management media backup
Use the following commands to back up the Digital Asset Management media uploads volume:
-
A command similar to the backup command outlined in Step 1 of Backup and restore a DAM image to create a backup of wp_profile can be used to back up the two
/opt/app/upload
and/etc/config
Digital Asset Management mount points.-
Refer to the following examples:
tar -cvpzf backupml.tar.gz --exclude=/backupml.tar.gz --one-file-system /opt/app/upload
-
tar -C/ -cvpzf backupmlcfg.tar.gz --exclude=/backupmlcfg.tar.gz --one-file-system etc/config/*
See Backup and restore a DAM image for more information.
Alternatively, the Kubernetes documentation pages present additional options to backup and clone persistent volumes.
-
- CSI Volume Cloning
Note
If either of the methods described in Step 2 as Volume Snapshots or Step 3 as CSI Volume Cloning is used, it is important to understand fuzzy backups with the wp_profile. A fuzzy backup is a copy of data files or directories that were operating in one state when the backup started, but in a different state by the time the backup completed. In case a volume snapshot or Container Storage Interface (CSI) volume cloning approach is used with the
wp_profile
, it is important that the snapshot is taken with the Digital Experience instance in shutdown state to ensure that recovery is performed.HCL Digital Experience has successfully tested the volume snapshot and CSI volume cloning methods with HCL Digital Experience 9.5 container deployments. It is recommended that customers perform the additional testing if they are using options Step 2 and Step 3 to manage the wp_profile backup.
-