BP job configuration using Data Domain as backup storage

[stewsie]

Hi

I have found a number of documents that discuss the configuration of the backup repository and job configuration when using DD. What I cannot find is any advice/recommendation about how the jobs should be created as far as OS type or application is concerned. Is it a good idea to have as few jobs as possible and if so is the recommendation to create jobs based on OS type? I need to configure the jobs to obtain the optimum dedupe performance. Our existing backup jobs are based on function so have a mix of OS type.

Any advice or recommendation will be well received.

Thanks

[Mildur]

Hi Paul

I want to note, that dedup appliances are not recommended as the primary backup target.
Restore speed is not the best and restoring is the main purpose of doing a backup.
Using a reFS or XFS disk repository gives you the same space savings with the benefit of really great performance (backup and restore). And they can be much cheaper than dedup appliances.

Back to the datadomain question.
For the job configuration, always use „Per-Machine Backup Files“. You can mix different OS or application, doesn‘t matter for the dedup rate. The datadomain will do global deduplication over all backup files.

Probably you know already about the other documents. But let me share them:
https://www.veeam.com/kb1745
https://bp.veeam.com/vbr/2_Design_Struc ... ation.html

[stewsie]

Thanks for the info and yes I am aware of your first point.

[Origin 2000]

I will not sign the first point... but anyways.

We use DD as primary target and when configure it veeam will set the options as "best as possible" during the setup. About the "Per-Machine Backup Files" which WE generally use it but i thought it needs to configure on target rather than job level. Dedup appliances dont like file changes/deletes so configure veeam to use separte sets of files for every VM is necessary. Keep notice that you need Enterprise edition or better... or VUL.

We only have one Main job which contains the complete mix of OS and APPs and 1-2 separate jobs because special time periods (The "largest" job starts first for example). The space savings we see o the DD are outstanding compared to the dozens of ReFS, XFS setups we also have.

With a entry level DD we see slow backup and way slower restore as with our previous backup appliance and compared to the ReFS, XFS setup.

Regards,
Joerg

[Andreas Neufert]

If you select a datadomain, the job settings around block size, compression and dedup are set automatically.

[stewsie]

I will not sign the first point... but anyways.

Thanks for the information. Very useful. How many VMs are in your main job? Thanks

[Origin 2000]

45-50VM are in the Mainjob.

Regards,
Joerg