dedup ratio always shows 1,0x

[elapuente]

Hello everyone,

We are backup up some virtual machines from a remote location (with a veeam proxy installed in that site). Since the beginning, we have a compression ratio of about 3,0x, but a dedupe ratio of 1,0x

I think that we should see some deduplication (1,0x means nothing if I am right)

Software is version 6.0 and totally patched

anyone with same issue?

Thanks for help and share knowledge

[Vitaliy S.]

Hello,

Can you please give us a bit more details about your job configuration? How many VMs do you have added to a single job? What is the OS version installed on these VMs?

Thank you!

[elapuente]

This is my configuration

All jobs incremental with synthetic full.
Daily backup.
Storage optimized for WAN target
Compression optimal
Enabled for inline deduplication
Enabled changed block tracking for vmware

Job01:

Windows 2003 Fileserver. 415GB
Windows 2008R2 DC - 30GB

Dedupe 1,0x (once 1,1x)
Compression: between 1,5x and 3,0x

Job02:
Windows 2008R2 Exchange 2010 - 150GB
Windows 2008R2 SQLServer + Fileserver - 145GB
Windows 2003 Navision (historic) - 40GB
Windows 2008R2 Xenapp - 50 GB
linux server (historic mail) - 70GB

Dedupe 1,0x (today 1,5x )
Compression: about 3,0x

After this numbers, I should change post title. It's always (except one time) 1,0x

¿should I make just one job with all machines?
¿is block tracking working so well than there is not enought data for deduplication?

Thank you for your support

[Gostev]

I would not expect to see any dedupe on the incremental backup files. Unless you do something like copying the same ISO file to all VMs in the job, in that case you should get nice dedupe. In fact, I did this very thing once without much thinking (was researching something else), and at first was quite puzzled with a very small incremental backup size... but then I remembered about our dedupe

[cffit]

[merged]

I setup about five different jobs for all my VM backups and grouped them by OS so I could take advantage of the deduplication part of Veeam B&R. However, now that I have the Enterprise Manager setup, the job summaries that it sends me say that the dedupe for every job I have running is only 1x. Is that common or is there a setting I need to change to get better dedupe numbers? I question if it is working since every job has the same dedupe ratio.

[Gostev]

What dedupe ratio do you get when you right-click the backup in the main UI, same? Just to make sure this is not some Enterprise Manager bug.

[cffit]

So in the end, if you are doing synthetic fulls and forever incrementals, there is no advantage to grouping backup jobs by OS or trying to match them up by common data? I guess I could undertand that, but it seems like Veeam really makes a big effort towards encouraging you to set your jobs up for best dedupe, which in most cases isn't much of an advantage if you are doing backups the way most people do.

[cffit]

When I look at the jobs in the main interface it usually says we get 99% or 100% on our incrementals and 59% or 60% on our synthetic fulls. I'm not sure what that exactly means.

[foggy]

Grouping VMs in jobs by the type of their guest OS you get a good deduplication ratio for the full job run (VBK file). Incremental changes are usually specific to each VM, while the full backups of the similar OS's contain a lot of common data that dedupe well.

[cffit]

What does a 60% dedupe number mean? Does that mean that it only had 60% of the data go through without dedupe and 40% was? Or how does that number figure in?

[foggy]

It means that the data was deduped to the 60% size of the original data.