Summary
In DPX 4.4.1, the archive to media feature is not supported for agentless backups on OSS servers. This article outlines how a user can perform tape backups of agentless backup data and then recover that data.
Resolution
An OSS installation can be comprised of two different types of client data protection technologies:
• A client with the DPX agent installed in the guest operating system; backups are performed to the OSS server through the agent (agent-based technology)
This article outlines how to backup and restore OSS data to tape without using the DPX archive to media feature, which is supported only for agent-based technology to OSS volumes without deduplication enabled.
To protect agentless backup data to tape, backup jobs must be defined using either File or Image backup technology. Both File and Image methods can be used for disaster recovery or individual VM/VMDK recovery. File backup allows the ability to recover individual files and folders, which is an inherent advantage to that technology. Image backup operates similar to the File backup method, but at the block level thereby providing optimized incremental backups. Which technology can be used depends on whether the OSS volume has deduplication enabled. If Windows 2012+ deduplication is enabled on an OSS target volume, File backup is not an option. In that case, it is required to backup and restore the volume in its entirety without file history using the Image backup feature. If there is no Windows deduplication, you can use either method, i.e. File or Image technology.
The document outline is as follows:
If Image backup is required, the procedure is simplified by protecting and recovering the entire volume from tape using that technology, rather than selectively restoring files.
Defining and running the backup job of OSS volume to tape
1. Define a backup job of an OSS target volume that you are protecting. This is accomplished by navigating to Backup > File and then drilling down and selecing the OSS target volume. You can select the entire server or individual volumes, but you should not select files/folders. (Note: a similar procedure is used for Backup > Image).
2. This procedure assumes that the agentless OSS data has not expired from the catalog. However, if the agentless backup to disk and OSS target volume backup to tape retentions are not equivalent, this will not always be the case. If the agentless backup data that needs to be recovered has expired and is not present in the current catalog, both the OSS target volume and the DPX catalog that refers to that data must be restored together in order to recover agentless backup data. Therefore, ensure a catalog backup to tape is scheduled to run periodically in conjunction with your recovery point objectives. Before performing a Catalog restore you should:
Defining and running the restore job of OSS volume from tape
1. The procedure here focuses on recovering files from a File backup to a new location. Subsequent steps show how you can use that data to selectively recover a VM/VMDK. If disaster recovery is the objective, simply recover the entire volume to the original location using Image or File restore. Then, proceed normally to recover data using Agentless VMWare restore jobs.
2. Navigate to Restore > File and select the volume you wish to recover data from.
3. Select “Set Browse Criteria”
4. Specifically select the recovery dates that encompass the base backup along with all subsequent incrementals leading up to the recovery point that you wish to restore. So, from the recovery point that you wish to restore, go backwards to the oldest data retained from the perspective of that snapshot. You can elect to go back farther in time than is necessary, but just make sure that the oldest, unexpired data at that time is included.
5. Press “OK” to “Set Browse Criteria”. Then, right click the volume and select “Restore Search”.
6. Enter the name of the backup job that contains the data you wish to recover surrounded by asterisks. In this example the job name is aCENTOS40_H so the search string would be “*aCENTOS40_H*”
7. Press “Search” from within the “Search Files to Restore” dialogue. The data that contains the search pattern appears.
8. Select all of the entries that include the base backup and all subsequent incrementals from the time period that you wish to recover.
9. Press “OK” and visualize that all of the items are selected for recovery.
10. On the right hand side, select a new volume on this or another OSS server as a destination that can be overwritten and used for the recovery. Do not restore the data to the original location.
11. Save the restore job without running it and then select “Preview Restore Job”. A window will pop-up indicating what tapes need to be made available to perform the restore.
12. After inserting the required tapes into the tape library, run the restore job to recover the data.
Defining and running the Agentless restore job from recovered OSS volume
1. Navigate to Restore > Agentless VMWare and run the “Agentless VMWare Restore Wizard”.
2. Simply follow the normal process for restoring data using the alternate secondary feature to specify the location you want to restore from as different from the original location. In the following window, deviate from the normal restore process by selecting the recovery point from the job to recover data that is included in your restore from tape. Then, select alternate secondary and point to the volume you recovered the tape data to, in this case the original data was on H: and we are recovering from the O: volume on the same server.
3. Answer the remaining prompts as if this were a normal recovery and run the job. The data will be restored as requested. The job log will clearly indicate the redirection of the restore to the volume that you specified via messaging.