I have to set up a structure like 7 daily, 4 weekly, 12 monthly and 5 yearly backups for my workplace, but am somewhat restricted in budget and storage space per device, so I have two NAS as backup repository, but each one allone is to small to save all the backups.
Now I am looking for a way to spread the backups over both NAS, e.g. save daily and monthly to NAS A, weekly and yearly to the NAS B. My initial idea would be:
A regular Backup Job for daily Backups, targeting NAS A, set up as
- Schedule: Run everyday at 9 pm; full backup every saturday
- Storage
- Retention Policy: 7 restore points
- Keep certain fulls longer: unchecked
- Configure secondary destinations: checked
- Secondary Target:
- Weekly Backup Copy Job
- Monthly Backup Copy Job
- Yearly Backup Copy Job
My next step is to set up the Copy Jobs, they all share the following settings:
- Immediate copy (mirroring)
- Object is the Daily Backup from above
- Schedule is any time
Depending on the backup the target differs:
- Repository is either NAS A or NAS B
- Retention Policy is always two restore points (can’t go lower) - which would be a duplicate of the newest two restore points of the daily backup, right? I don’t need them (I know, I should have a Backup at a different site/ cloud, but as said, no way my boss pays for that)
- Keep certain backups longer: is set to 4 weekly/ 12 monthly/ 5 yearly
The last two settings are what troubles me the most.
The GFS settings offer me a weekly option “create weekly full on this day”, which I set to saturday for Weekly and leave blank for Monthly/ Yearly, as that is the day the full backup is created for the Daily.
But the monthly option refers to the weekly Backup and asks explicitly for a weekly backup to keep longer - can I just check this box, set it to e.g. “Last” and tell it to keep a dozend Monthly, allthough the weekly Copy is not configured for this specific Copy Job? If I understood the flagging system correct, Veeam should recognize that it has to only create one Copy and flag it weekly and monthly and just delete those flags after different times, thus keeping some Backups (the last weekly of every month) longer, is that correct?
Second question: can I really set the retention policy to the minimum for my backup copies and to seven restore points for the Daily? Will this create a problem with the Copies like the source Backup having allready been purged when the longer cicle of the Copy Jobs triggers?
Of course, if the whole setup is going to fail I would be happy to get poked in the right direction to setup GFS over multiple storages. Or if that is impossible at all, at least that would be good to know, and if there is no way to do so, I will have to let my boss decide which risk he want’s to take, but at least I could explain it to him confidently before realising the backups did not work as planned after disaster strikes...