Is this backup strategy possible with Veeam Replication ?

[SomeDude]

Hey Guys,

I'm currently planning a new backup concept for our company (we already bought Veeam )

My current strategy is as follows:

daily backups: 7 days retention (1 full at monday, 6 increments tues - sun) to disk
weekly backups: 4 weeks retention (1 full at first week of month + 3 increments on weeks 2-4) to tape
quarterly backups: 4 quarters retention ( 1 full after 3 months, incremental after 6 months, full after 9 mo, incr. after 12 mo) to tape
monthly backups: 12 months retention (1 full at month 1, increments at month 2, 3, full at 4 and so on, adding up to 12 restore points with 4 fulls) to tape
yearly backups: 10 years retention (10 fulls at every start of the year) to tape; we want to remove every full year from the library and store it someplace else. So I think we need a dedicated tape pool for this

Is this concept possibe with Veeam ?

I'm currently thinking to realize this with a single GFS-pool containing a load of tapes, but I'm not sure if this is the best practise.

[PTide]

Hi

I'm currently thinking to realize this with a single GFS-pool containing a load of tapes, but I'm not sure if this is the best practise.

In your case it won't work since GFS pools cannot be used for incremental backups.

Daily backups: 7 days retention (1 full at monday, 6 increments tues - sun) to disk

Yes, it is possible.

weekly backups: 4 weeks retention (1 full at first week of month + 3 increments on weeks 2-4) to tape

Not directly, but you can create a Backup Copy Job to create a restore point every week, and create another backup to tape job that would grab a new restore point every time the restore point appears, and close weekly media set after the 4th week incremental has been copied.

quarterly backups: 4 quarters retention ( 1 full after 3 months, incremental after 6 months, full after 9 mo, incr. after 12 mo) to tape
monthly backups: 12 months retention (1 full at month 1, increments at month 2, 3, full at 4 and so on, adding up to 12 restore points with 4 fulls) to tape
yearly backups: 10 years retention (10 fulls at every start of the year) to tape; we want to remove every full year from the library and store it someplace else. So I think we need a dedicated tape pool for this

Pretty much in the same way as I've described above - you'll need a supplementary backup or backup copy job to produce a source chain of a desired configuration so a tape job could copy it to a media pool.

Thanks

[SomeDude]

In your case it won't work since GFS pools cannot be used for incremental backups.

Thanks, but wouldn't it be easier if I create 4 simple Media Pools (one for weekly, monthly, quarterly, yearly) and set the retention period accordingly ?

Or is the best practice for this concept the way you already mentioned ?

[PTide]

Thanks, but wouldn't it be easier if I create 4 simple Media Pools (one for weekly, monthly, quarterly, yearly) and set the retention period accordingly ?

Sorry, I'm not sure what you mean by "easier" - GFS pools simply do not allow you to send incremental backups to them, so there is no other way than to have simple media pools with different retention settings, if needed.

Thanks

[SomeDude]

I think we both meant the same,sorry for the confusion.

If I want to use GFS pools and assume 40 TB of backup data, do I need:

weekly: 4 * 40 = 160 TB
monthly: 12 * 40 = 480 TB
quarterly: 4 * 40 = 160 TB
yearly: 10 * 40 = 400 TB

1200 TB of space just to backup the data without any growing ? (Uncompressed data).

https://helpcenter.veeam.com/docs/backu ... tml?ver=95 quotes GFS with

It lets you keep backups of machines for an entire year and requires minimum amount of storage space

This doesn't seem to match since "only fullbackups" excludes minimum amount of storage.

[PTide]

First of all, the link that you posted is about Backup Copy Job GFS, which is different things. A tape job do not archive the GFS restore points from disk.

Secondly, GFS retention does not duplicate restore points, if you set the same day for weekly/monthly/yearly, the backups will not be duplicated, but the same file will be marked as Weekly and Monthly. Please check this "Restore point simulator".

Thank you

[SomeDude]

Thanks, I've done some calculation for the GFS pool option (and got some questions):

Our current total amount of Data is ~86 TB.



1. Why do I need to specifiy Retention Points if I specify those after "Backup Copy Job Specific" ?
2. Whats the meaning of "Work Space" ?
3. /Edit: Ribbon is totally different than right clicking, got this now.

P.S. You said earlier that I need to create a Backup Copy Job for monthly etc. backups, but the maximum copy interval is 100 days, how does this work if I need 365 days i.e. ? Did you mean to create a Copy Job?

[PTide]

1. Because that is the number of restore points you'd like to keep in your simulated Backup Copy Job chain. Without that number it would be impossible to calculate the required amount of space. If that does not answer your question, then please elaborate

2. As documentation says:

The output shows the result after the job has run. However during the backup, the job might temporarily need extra space because the retention policy is only executed after backup. Also, in some case, it might be required to run an active full manually, which will take extra space. Thus the working space is in correlation with this temporary space.

So, basically, that's the amount of space you should have for maintenance tasks.

I meant Backup Copy Job. The copy interval determines how often a backup copy job will transfer new restore points from a source repository. Please check this User Guide page. The number of restore points to keep in your backup copy chain is determined by "Restore points to keep" variable on the "Target" step.

Thanks