Windows 2019 File Cluster with Storage Replica

[logixcom]

We have a 6 node Windows 2019 File Cluster with 2 roles providing file shares/disks for users. Role A is for App related file Shares and Role B is for User related files shares/disks. # Nodes are in Datacenter A and 3 nodes in Datacenter B. We host the active shares in DC A and do Windows storage replica to DC B, it works quite well. Total data size is about 13TB combined. I should also mention the disks are ReFS with dedupe on, so data is deduped down to about 6TB. We are using Veeam Agent with the Cluster added per Veeam best practice as each Node is in VMWare using RDM disks. Veeam is backing up the data as intended, but when I look at the actual backup file size it is 30+TB. I have opened a support case and so far they are unable to provide a reason as to why the backup file is double in size. Anyone doing a similar type backup and are you seeing the same results?

We also do a tape backup of the latest copy and it is "chewing" through tapes so would like to either fix this or get an understanding of why this is happening so we can budget for more tapes.

We have licensing available to switch to File Share backups if need be, but I am hoping not to have to "eat up" those licenses. We are currently using version 11 but are in planning to update to version 12.

Any insight would be greatly appreciated.

[HannesK]

Hello,
and welcome to the forums.

Can you please post the case number that we can check what support already investigated?

In general: V11 is bloating agent backups in various situations (backup copy jobs, virtual synthetic fulls to tape, capacity tier). That is fixed in V12. But that does not match the description of your post, so I guess V12 does not help here.

Best regards,
Hannes

[logixcom]

Veeam Support - Case # 06128558 - Windows File Cluster Storage Replica double data

[logixcom]

You seem to be facing a known behavior where, during synthetic operations, such as Synthetic Full backup, merge, or cloning of cluster disks between nodes inside the backup file, the deduplication index limit for data blocks can be reached, causing the backup file to become bloated
To mitigate this, please follow the instructions below.

1. Please edit the job and Increase the size of the data blocks to 4 MB, inside the backup job settings.
-- And then --
1. On your repository server, please create the registry key below. Please note that if have other cluster jobs pointing to other repository server, the registry key must be created on the other repository servers as well.

Location: HKEY_LOCAL_MACHINE\SOFTWARE\Veeam\Veeam Backup and Replication\
Name: DedupeIndexLimit
Type: REG_DWORD
Value Decimal: 25165824

[logixcom]

Issue has been resolved. Backups are now the correct size for the data it is backing up.