Veeam with blob storage and caching disk

[zic04]

Hello,

I have the following problem :

I want to send backups in azure with Veeam 10.
Veeam Backup and réplication installed in the servers on-prem.

In backup Repositories I added the blob storage azure.
I've also added a local drive to realize a cache backup for Azure.

The backups are working, I find the same backup on azure and on the disk.
The caching disk on the other hand does not erase the data, I do not want to keep a single day on the disk.

Safeguard Plan :

Local datastore :
Every day during 31 days

Copy datastore :

This for datastore Azure
Weekly backup : 4
Monthly Backup : 3
Quaterly Backup : 4
Yearly backup : 10

In the caching disk we want only 1 day it's possible or exactly the same.

thank you for your help
Patrice

[oleg.feoktistov]

Hi Patrice and Welcome to the Community Forums!

As far as I understood, as per your configuration, you want 1 restore point on your local SOBR extent and only GFS backups to be offloaded to Azure, correct?

Thanks,
Oleg

[zic04]

Hello,
Yes :


on the local backup I want to have 31 days : it's ok


Copy backup job : "scale out repository


Datastore with disk 1 days
On Azure 12 month / 10 years

Thanks
Patrice

[zic04]

if you want to see the configuration :

https://we.tl/t-KBIpHgWxRS

Thanks

[oleg.feoktistov]

As far as I see from the screenshots, you're using Backup Copy job with mirror mode. So, keep in mind the following:
- Only full backups with GFS flags, which belong to inactive backup chain are moved to capacity tier.
- It is much more dependant on source backup job schedule than Backup Copy with periodic mode.
That's why you see the behaviour like that.
If you want to use Backup Copy job to keep one full backup on performance tier and GFS on capacity, the other way around would be
to setup Periodic Backup Copy job and specify its interval corresponding to the time each full is created with a source job. For example:
- With a source backup job fulls are created weekly every Wednesday.
- Interval for Backup Copy job is set to 7 days starting from Wednesday.
But I'd recommend to stick to the chance to copy increments.
Thanks!

[oleg.feoktistov]

What is your source backup job setup anyway? Maybe it'd be more relevant for you to just offload GFS backups in scope of primary job.

[zic04]

Hello,

Thanks for your information, I understand.
I still have two more questions.

1) When files are sent on azure there are still files left for consolidation. If we lose the disk, it's impossible to get the data back down ?

2) if we take the example of a 1 vm : Name : Filer Size 4To

In local we want 31 Days with synthetic Full every 7 days

In Azure we want 12 month and 10 yearth

Do you have any idea how much space on the disc for scale out repositories ?

thank you for your precious help

Regards,
Patrice

[oleg.feoktistov]

Hi Patrice,

1. Right, for those files moved to capacity tier only metadata is left on performance extents. However, it is also copied to capacity and placed under /storages directory. So that, if your performance extents are unavailable, you can still appeal to a metadata in your object storage to act on offloaded blocks. You might want to take a look at the examples on how data is moved and Restore Scenarios for such cases.

2. That would depend on several factors, like Compression and Deduplication ratios or whether or not your repository uses block cloning.
You can begin with checking out these tools:
- Restore Point Simulator can help with size and space calculations.
- Veeam Size Estimator can assist with workloads design.

Thanks,
Oleg

[zic04]

Hi,

Thank you so much.

Patrice