Basic backup policy

[t_a]

Hi,
I've already gone through the manual, read the forum and by trial and error I was unable to set the execution and maintenance of backup copies according to my rules.
The first plan I would like to get.
1) Hyper-V Backup. My assumptions are:
Full Backup -> 7 Restore Points -> Full Backup.
During the second full backup, 7 restore points are deleted and only the full backup is left.
2) Windows Agent Backup: Full backup -> 7 restore points->Full backup.
During the second full backup, 7 restore points are deleted and only the full backup is left.
Below are my settings and I don't know why I already have 9 restore points and the program did not remove the restore points when making a second full backup.
How do I set Job to work according to my rules?
https://imgur.com/a/K6LiPpZ

[Gostev]

Hi, just think about that: if it removed the previous full backup chain now, then you would have been left with 2 restore points instead of the desired 7 restore points. Veeam will be able to remove the previous full backup chain when the last incremental backup in the chain is no longer under your retention policy.

[t_a]

Thank you for your reply.
However, according to my assumption, I no longer need intermediate points and a full backup from 13/05/2023 to 19/05/2023, because on 20/05/2023 the program made a new full backup that contains data from the previous week. I have limited disk resources on the NAS, so I would like it to automatically remove old points and a full backup 13-19/05/2023.

[Gostev]

If you really want a backup created on 19/05/2023 removed on 20/05/2023, then you should set your retention policy to 1 restore point instead of 7 like you have it right now.

This would be a terrible way to do backups though, as you won't be able to rollback if you don't spot a data corruption or malware in a timely manner. I mean, there will be days where you will be sitting with a single most recent restore point. And if a data loss event happens just before the corresponding backup, then this is what will be backed up and that single restore point will be completely useless.

[karsten123]

Hint: Forever forward incremental keeps exactly the amount of restore points you configured in the job. As NAS SMB repositories are not capable of leveraging fast clone, it is the best way to go (in most cases).
Thoughts?

[Gostev]

Not to use SMB NAS as repositories. They should never be considered as backup targets for reliability reasons.

[karsten123]

Of course your right, Anton. But if it is there allready. Sadly, reality is not often ideal.
I would never recommend a SMB NAS as a repository, too.

[t_a]

Thanks for the answers.
When it comes to NAS, it is a RAID 6 array. When it comes to protection against malware, QNAP takes snapshot copies and maintains them for 7 days.
I think it's a pretty good protection against physical disk failure and malware.

That's why my question was about: How to set the backup policy in the program according to my assumptions, and not to develop the storage method.
Can anyone help me and tell me how to set the schedule?

[Gostev]

RAID will not protect you against storage-level corruption, in fact software-based RAIDs *are* the reason for most of them. These corruptions are the very reason why enterprise-grade RAID controllers come with battery-backed write cache.

Storage snapshots are not immutable or air-gapped and will be deleted along with the production volume by a hacker, so don't count on them in case of a proper cyberattack.

I've already explained above how exactly you should change your retention policy to achieve your desired backup removal behavior per you example.

But let me be crystal clear: you're setting yourself up for a recovery failure with literally every single part of your plan. Too many people with similar ideas have failed to recover their data, which is why this community is so passionate about trying to prevent others from repeating the same well known mistakes and bad practices.