Help/Advice to sort out my growing backups

[mjgreen81]

Hi all,

Firstly, a few key point about our configuration:

We are running Veeam 9.0.0.1715.
We have approximately 30 backup jobs, backing up a total of approximately 47 VM's.
All jobs are configured as reverse incremental (with varying number of restore points)
None of the jobs have "Exclude deleted file blocks" selected
None of the jobs have Active Full Backups configured
We backup to a SAN which provides approximately 5.3TB of storage. We are getting to a point where we only have around 300GB of storage left.

So everything has been running smoothly, but I am getting low on storage space. I have started to do some research today and I have discovered that I might be able to use deduplication (not something I was previously aware of) and the "Exclude deleted file blocks" (also something which I previously was unaware of) to claw back from storage.

I am looking to get some advice on deduplication and how effective it is.
Is it just a case of re-configuring my backup jobs so that VM's which were built from the same template are in the same job and this cuts down on the duplication of data?
What about if they were not built from the same template but are the same OS? Does this cut down on the amount of data that is backed up?
Is there a way to roughly estimate how much storage might be saved? (For example if I had 2 x 60gb servers built from the same template which were previously in 2 separate jobs, but were now in the same job how much storage would I expect to see recouped).

I have also been reading about the "Exclude deleted file blocks" option, which I have not noticed in previous version of Veeam. It sounds like this could save me some storage (especially on file servers etc) I did some reading up about this and believe that in order for me to take advantage of this, I need to periodically run a Full Active Backup (I subsequently understand that this is probably something I should be doing anyway to avoid corruption in the .VBK).

So my questions regarding the Full Active Backup are:

Should I enable it?
Will it mean that I have a second full size .vbk alongside my .vbk from the reverse incremental? (I am going to struggle to accommodate this due to lack of storage)

Are there any other tips to help me manage my situation with consideration that I only have around 300Gb spare to play with at present?

Thanks in advance!

[lando_uk]

You really need to reduce the number of jobs. If you had more VM's in each job, they will deduplicate against each other saving you loads of space. For 47 VM's you could reduced them down to 2 or 3 jobs.

[mjgreen81]

Thanks for the input - I had not realised the impact that having so many multiple jobs would have.

I will have to do some planning and work this one out.

Also, I ran a full active backup on one of the jobs and the .VBK file dropped from 67GB to 47GB, so there is definitely some work to be done here.

[mjgreen81]

I also just repeated the above, this time selecting the "Exclude Deleted File Blocks" option and it has dropped down again to 12GB!

I will have to do some thinking to plan all this out, but it looks like I could make claw back some significant space with some tweaking.

[JoshuaPostSAMC]

What kind of speed are you getting with Reverse Incremental? I was getting 5-10 MB/s with my Reverse Incremental job and recently changed it to Forward Incremental and now I'm getting 80 MB/s on the exact same job. Backup window down from 5 hours to 30 minutes. Down with Reverse Incremental!

[mjgreen81]

Backup speeds are fine, I get well over 80MB/s using reverse incremental.

[mjgreen81]

Just to give some people some real world statistics (in my case anyway), after I grouped together 30 x VM's into one job (to take advantage of deduplication) and also selected the "Exclude Deleted FIle Blocks" option, these were my storage statistics:

Previously these 30 VM's were split across 10 x jobs (some with 1 VM, some with up to 5 VM's).

In total, all the .VBK's added together = 1255GB
After grouping them together and selecting Exclude Deleted File Blocks, the size was 601GB. Over 50% smaller.

[lando_uk]

Good stuff. Some of my jobs have grown too big now with 8TB+ VBK's - They take forever to transform and health check, keeping them under 4TB I think is a good idea.

[mjgreen81]

8TB! wow!

[lando_uk]

Every change has an effect though. Your VMs probably restored quicker when they were in single jobs. All the space saving features generally have a negative effect on backup and restore performance.

[mjgreen81]

Thanks for the tip. Its extremely rare that I would have to restore anything from any of the VM's which are contained in this particular job.

When I move on to creating new job's for the VM's where I am much more likely to need to restore files (e.g. file servers) I will think about it more carefully and probably have less VM's per job.

[foggy]

So my questions regarding the Full Active Backup are:

Should I enable it?
Will it mean that I have a second full size .vbk alongside my .vbk from the reverse incremental? (I am going to struggle to accommodate this due to lack of storage)

Yes, every new full will start backup chain from scratch. Please see this thread for some considerations regarding the need for active fulls.