Retention

[ftbug]

Hi, we are currently backing up a customers data who have asked for the following rules;

Hourly restore points over the last 24 hours period (24)
Daily restore points for the last 1 week period (7)
Weekly restore points for the last 1 month (4)
Monthly restore points for a year (12)
1 Yearly restore point kept for the last 7 years (7)

We have currently specified this in the properties for the Backup Copy job, however, this seems to be taking up a huge amount of space on our storage.

Is there any other better way to specify these particular retention rules?

[Shestakov]

Have you set the retention with GFS option?
The way to save space for you is to set the same days for weekly/monthly/yearly etc.
E.g. if you set weekly on Sunday, monthly on the first(last) Sunday of month and yearly on some Sunday as well, the only .vbk file will be marked as weekly/monthly/yearly and space is to be saved.
Thanks!

[ftbug]

Hi, thanks for the reply.

We are currently using the GFS option and all restore points are scheduled on Sundays.

It seems strange that the customers data is siting around 40TB when the disks being backed up total around 8TB.

[skrause]

With that number of restore points, it makes sense your backups will take up more space than the production VMs.

GFS, by its nature, has to make a Full (synthetic) for each restore point saved - you need to be able to restore to that point without having to pull an older full + increments.

Are you keeping all of your monthly/yearly's on disk?

[Shestakov]

ftbug, as Steve mentioned, you have several .vbks of each VM, so it`s fare that total size of those files bigger than the source data.
Could you provide the average size of full backup file?
Thanks!

[widmerkarl]

Hi ftbug,

This is a good kind of retention. You can cover nearly every "i deleted my word document" case or also bigger issues like corrupted server after a crash or installed patches etc.

It would be good to know what's the average size of a full backup. So you can calculate the needed amount of storage space.

And i'm just curious how that infrastructure looks like. I know, it doesn't have something to do with the thread. I'm just wondering how you can do hourly backups of all VM's. When i compare with my customers, the biggest infrastructure size is about 4 hosts, shared storage and about 40 VM's, round about 5 TB total file size. I have no chance to do hourly backups.

[Shestakov]

By the way there is a tool to estimate the size of restore points.
Just don`t forget to check Backup copy w/ GFS in the first drop-down menu.