help with scheduling/retention

[Ioannis.T]

Hello,
I need some insights/suggestions regarding our scheduling and retention setup.

We have around 250 servers that account for ~100TB which in turn are around ~50TB when backed up. (compression is doing nice job ).

Current setup:
1x Backup server with a Tape library:
Schedule: Weekly full backups with forward incremental
Retention period:
- each Weekly and each incremental we keep for 12 weeks
- Last backup of each month we keep for 12 months

At the new setup we have 2x linux repositories each of 180TB. The current tape library can be used but we would prefer to avoid it if possible.

We were thinking of doing a reverse incremental for 90 restore points, and hence cover the 12 weeks retention we had so far. That would need almost one of the repos.
I am not sure how to do the "keep the last backup of each month" for 12 months. The other repo is not enough so we need to use the Tape library for this part.
Would it be possible to break the same jobs to different repos depending on the retention period? so I can then create a tape job to move these backups to the tape?

Of course the way of thinking might be wrong and a new approach might be better, so please feel free to suggest any ideas and approaches.

Thank you !!

[chris.childerhose]

One suggestion would be to use the Linux repos and set the GFS options within the backup job to accomplish what you want to do. Reverse Incremental is not recommended as an option with backups anymore and in a future version will be deprecated (removed). The easiest way is GFS though - https://helpcenter.veeam.com/docs/backu ... ml?ver=110

[Ioannis.T]

Hello Chris,
Thanks for replying.
Is there a way to separate the last monthly backup to go to a separate repository instead of the configured one for the rest of the days/weeks of the month?

Thanks

[Ioannis.T]

To clarify,
We are looking to keep the one repository for our 3 month retention period, but redirect the last backup of each month to another repository to keep for 12 months.
We cannot seem to be able to find a way to do this.

[foggy]

Not within the same job, but you can create another one pointing to another repository.

[chris.childerhose]

Yes as noted by @foggy you need to create another job for using two repos.

[Ioannis.T]

bonus question
is there a way to do Bi-weekly Full backups?
I can only seem to be bale to achieve either weekly or Monthly

[chris.childerhose]

I think you would need to possibly create a second job and then change the scheduling of it to be the opposite weeks of the other job.

[foggy]

Or use some scripting to enable full backup in the job settings on the required days, run the job and disable full back.

[chris.childerhose]

Another great idea. Thanks for that one foggy.

[darren.flatman]

Not sure if its an option but you could use a scale out repository

Performance would be two Linux repo
Capacity could use something like AWS S3
Archive would then need to be AWS Glacier storage (which is pretty cheap)

Set the limit to 80 or 90% max on performance then move to capacity or do by date
Then move all Monthly GFS backups to Archive
Capacity Tier can be made immutable for 365 days and Archive Tier indefinitely should cover whatever you need Tape for and you don't have the hassle of managing a tape drive

[david.domask]

While the above ideas aren't bad, do I get it that the goal is:

>Would it be possible to break the same jobs to different repos depending on the retention period? so I can then create a tape job to move these backups to the tape?

If so, use Backup Copy for this. Your Primary backup handles the short term retention, and your Backup Copy handles the GFS retention on the desired repository. Tape-out from the Backup Copy or the Primary, dealer's choice.

1. One repo for shorter term retention and faster restores, maybe configure weekly GFS if you like
2. Set up BCJ with Monthly GFS
3. Configure Tape-out from the desired repository (fastest is best so you can serve up data to the tape hardware as fast as possible)

Would this maybe work for you? With XFS on those nice new linux boxes, you also get space savings on the GFS points, so you extend your storage usage quite a bit.

[Ioannis.T]

Hello,

Thank you all for the responses.
Very useful input.

What I am failing to understand is how to achieve the 2-weeks of retention on the "primary" repository.
So the Full backup can happen every 2 weeks.

What I see that Veeam fails to do is the following.
Although it has the option to set a full backup every first/last,/second/third/fourcth Sunday of the month, there is not option to set it as multiple options. for example I want full backup every 2nd and fourth Sunday.
The other thing that I cannot find is to exclude days. In the scheduling part of a job you can set it to run at the 12 day of each month for example. If at that point there was an option to select multiple days it would allow you to set up which days of each month you want a backup to run.

I know that I might repeat mylf , but I am really trying to understand the scheduling part and how to use it as to my needs.

For one more time, thank you all for the responses.

[foggy]

You can use an external scheduler (Windows Scheduler, for ex.) to trigger the job via script on the required days.

[chris.childerhose]

As per @foggy as there is no built-in every 2 weeks option. This is the option to use.

[Ioannis.T]

thank you again.
I will return with what we decided to go with