Basic backup scheme setup

[ilovepancakes]

I am new to Veeam and backup schemes. Looking to implement the following backup strategy but would take suggestions on better ideas too. What is the best combination of settings and jobs in order to make the following happen?


1st: hourly backups for the past 72 hours.

2nd: daily backups for past month.

3rd: monthly backups after that for 3 months.

[Dima P.]

Hello and welcome to the community Chris!

May I ask what type of data are you planning to protect (VMs, physical machines or maybe files to tape media)? Thank you in advance.

[ilovepancakes]

Thank you. Just VMs for now. It would also be great to implement S3 object storage into the above setup for the daily and monthly backups so they stay safe on the cloud.

[foggy]

You can create a regular backup job with hourly schedule that will keep 72 restore points and the backup copy job with daily copy interval that will use backups created by the first job to keep 30 daily restore points in a secondary repository. In the backup copy job enable GFS retention to keep 4 monthly points. For the cloud backups, consider using scale-out backup repository and its Capacity Tier.

[ilovepancakes]

OK, I think I know how to set that all up, however my assumption was to create the regular backup job, telling it to store the backups on a local drive, then tell the backup copy job to use the S3 object storage as it's storage method. 2 issues though...

1.) If I have the local drive added to a created SOBR, it won't show up as a destination for the regular backup copy job. Only the SOBR shows....

2.) I only want the S3 storage to be used for the backup copy job (daily and monthly backups), so even if the backup copy job has the SOBR as destination and the system creates the copy backups daily and monthly on local then offloads to S3, the hourly backups from above would get offloaded to the SOBR too from above, right?

I guess essentially to do what you suggest I have to be able to select the local drive as direct destination for regular job and then the SOBR as the backup copy job destination?

[foggy]

After being added to SOBR as an extent, a backup repository cannot be used as a simple repository anymore. Also, backup copy cannot use the same repository both as the source and the target one. You can create another simple repository in adjacent folder on the same disk though.

I guess essentially to do what you suggest I have to be able to select the local drive as direct destination for regular job and then the SOBR as the backup copy job destination?

Correct.

[ilovepancakes]

OK great, that is what I did then. I have the regular job doing incremental only (no synthetic or active fulls, is that right?) with 72 restore points and it runs every hour. Then I have the copy job running to the SOBR daily with 30 restore points and the option to keep 4 months of monthly backups.

Also, I have the SOBR set to offload objects older than 0 days and gave it permission to run anytime of the day or night on every day of the week. But, last night even after the backup copy job ran it's first time, the SOBR doesn't seem to have uploaded anything to S3...

[foggy]

Please review this section for details on how Capacity Tier works, should answer your question.

[ilovepancakes]

OK I already had it configured as mentioned in that section of KB for keeping monthly restore points for archival purposes and assuming nothing got offloaded yet because I didn't hit the first Sunday of a month which is the schedule to create the monthly full for archival purposes in the backup copy job. This is fine.

BUT, I am wanting to have it offload daily backups created by the backup copy job at 1:30am every morning. So the backup copy job is set to copy every 1 day at 1:30am with 30 restore points. But, it is making incremental backups each day at 1:30am which is why it seems the chain is not being "reset" and the daily copies offloaded. How can I get the backup copy job to make daily synthetic fulls from the hourly backups being made by the regular job, or, just make active fulls at 1:30am and offload those?

I was assuming the way backup copy works is it continuously copies each hourly backup as it is made by regular job and then writes them into a full backup file at 1:30am each day.

[foggy]

First, backup copy works differently, it doesn't copy backup files from the source, it synthetically creates a single restore point per interval copying the latest VM state available on source. Second, only full backup files that have a GFS flag on them are offloaded to Capacity Tier, incremental backup files are not offloaded.

[ilovepancakes]

OK, so in order for the backup copy job to create a full backup, which is what is needed for the offload to happen, the regular job must be set to create periodic full backups?

[foggy]

No, regular backup job settings do not affect backup copy jobs. GFS restore points (which are full backups each) should be already created, since you've enabled the option to keep 4 monthly backups.

[ilovepancakes]

Right, that's what I am asking about then.... how do I get daily GFS restore points in the backup copy job? I enabled my option for 4 monthly GFS points but where do I tell it to do daily GFS points/full backups?

[veremin]

You cannot have daily GFS backups - the most frequent schedule for GFS restore point is weekly.

However, you can make backup copy job produce active full backup on daily basis, using PowerShell. (Start-VBRJob cmdlet)

Thanks!

[ilovepancakes]

Alright, but if I use that command to schedule active full backups made daily as part of the backup copy job, it still won't upload them to S3 as part of an offload, or will it?

[foggy]

No, only GFS restore points are offloaded.