Questions About Backup Copy Behavior with ExaGrid

[rin]

■Configuration:
Veeam Backup Server (1): Windows Server 2019
Veeam Proxy & Repository Server (1): Windows Server 2019
Veeam Version: Backup & Replication 12.1.0.2131
ExaGrid Version: 7.0.0.P04 (build 422)

■Current Situation:
We are performing backup copy jobs from the repository server to ExaGrid using the forever forward incremental method.
Recently, we’ve noticed that the time required to complete these jobs has increased significantly after reaching the maximum number of retention points.
To address this, we are considering applying the best practices described in the following article:
https://www.veeam.com/kb1745

■Concerns:
As the backup copy method will change from forever forward incremental to forward incremental, we would like to confirm the following points regarding the impact on data stored in ExaGrid:
1. After switching the job type, will synthetic operations be triggered during every backup copy job execution until the existing backup chains are removed?
2. Will the existing backup chains be deleted immediately after the configuration change?
3. Are there any behavioral differences between backup copy jobs targeting ExaGrid and those targeting a standard repository?
4. Are there any other potential impacts we should be aware of when changing the backup copy job configuration?

We would appreciate your guidance and clarification on these points before proceeding.

Best regards.

[david.domask]

H rin,

Will go through the questions one by one:

> 1. After switching the job type, will synthetic operations be triggered during every backup copy job execution until the existing backup chains are removed?

Synthetic operations happen during GFS creation, so you will see Synthetic Operations during those days. Exagrid supports Fast Clone so these will be pretty light-weight operations if using Fast Clone. Else since it's Backup Copy, there may be value in the Active Full method which avoids Synthetic Fulls entirely by reading all data from the source repository.

> 2. Will the existing backup chains be deleted immediately after the configuration change?

No, the chains will continue. Switching to Forward Incremental simply tells Veeam to create periodic Full backups (Synthetic or Active Fulls), which can be done without disrupting the current chain


> 3. Are there any behavioral differences between backup copy jobs targeting ExaGrid and those targeting a standard repository?

As Exagrid is a deduplicating appliance with a landing zone that is not deduped, functionally not too much difference. Follow the KB article you identified and follow the Exagrid best practices from the Vendor Specific Advice section.

> 4. Are there any other potential impacts we should be aware of when changing the backup copy job configuration?

Nothing too much except that there will be periodic full backups. As recommended above, if your Exagrid version supports fast clone, it's recommended to use that for fast and lightweight (read: space-less) Synthetic Fulls.

I'm going to guess with your current setup, the Forever Forward Incremental merge was taking the longest hence the interest. Indeed for deduplication appliances Forever Forward Incremental is not recommended for this very reason. See how the jobs run with the KB recommendations, it should be smoother experience.