VBR Backup report counts empty space as "dedupe"

[hyvokar]

Hi,

just noticed.

Created a new backup with Veeam Backup agent (windows). When I check the backup report from VBR, I get 3.6x dedupe ration (wow!). Then again, when I look at the data read and data transferred, 6.5TB and 6.2TB, so it seems that dedupe calculation counts into also the empty space in disk (16TB) in . Is this by design?

[HannesK]

Hello,
I cannot confirm what you see in my lab. Do you maybe have a support case number including screenshots and more details (version number, log files, was it full or incremental backup, managed by server / managed by agent) ?

Best regards,
Hannes

[hyvokar]

Hi,

haven't open support ticket, since I'm using community edition and support team based on past experiences has their hands full on paying customers.

Veeam agent is installed on windows workstation, 5.0.3.4708, VBR is installed on Hyper-V virtual machine, ver 11.0.1.1261

Nothing special in processing or in the logs. Job is succesfull. Tried two different dedupe options (WAN target first, then local target), same results in report.

https://ibb.co/xLGjRRj

[jorgedlcruz]

Hello,
Thank you for the report, always love to see a good HTML report in green out there.
So, the deduplication field is calculated looking at the backup fnal size (6.4 on your case) in relation to the total data size (23.3 TB), of course, considering that swap file blocks are still excluded, bitlooker and others. That number will always be huge. The dedupe option you select as well influences on how we create our backup file as well, meaning blocks of 1MB, 4MB, etc. But that is relevant if you are gonna do something else later on with the file, meaning sending to a dedupe appliance, or use Veeam Capacity Tier to send it to Object Storage, etc.

Compression is just 1x as you saw, you can always try to increase the compression on the job at the expense of more CPU consumption.

If the backup file, and the future chain meaning your incrementals and synthetic are taking too much space. You can always put them to a Repository where you enable Windows Deduplication.

Hope this answers your question.
Hope the explanation makes sense to you.

[hyvokar]

Would it make more sense to calculate deduplication ratio compared to actual data read (not counting the empty space) ?