Suggestions for long-term file versioning?

[stamm]

Hi everybody,

I'm wondering if there are any suggestions for a problem I'm facing. Due to the nature of our business (architecture), my company keeps a monthly snapshot of project data on tape for seven years for archival and legal purposes, essentially maintaining 84 restore points. We've been copying monthly full backups to tape and sending them offsite for many years. The goal is to get away from tape, and we would like to use Veeam B&R with a Cloud Connect repository to house future backup copies without breaking the bank. One issue is that our active projects total about 18TB right now so we'll need a lot of cloud space.

Currently our Veeam environment pushes backups to in-house Exagrids, and active fulls are run every two weeks with incrementals every day in between. Is switching to forever incrementals likely the answer? That would get me 84 restore points after seven years, but how would I eventually delete data older than 84 months? And I also don't have a realistic estimate of how fast the Exagrids will run out of space with forever incrementals.

I'm hopeful that someone else has been through this already.

Thanks!
stamm

[dellock6]

Pat,
I believe the answer is not in extending or modyfing the retention of your backups, and changing backup mode would not make any difference.
Have you looked at Bacup Copy jobs instead?

https://helpcenter.veeam.com/backup/vsp ... _copy.html

With them you can keep your primary backup jobs unchanged, and leverage backup copy to a secondary location (like Cloud Connect) and use for example the GFS rotation to extend it to multiple years without dealing with a huge number of restore points.

[stamm]

Thanks Luca,

I do intend to use backup copies, but the issue with doing that with full backups is that I'll eat up a minimum of 18TB of cloud space every month. Our data usage always goes up and never down. That cost will rapidly spin out of control with cloud storage, which is why I threw the question out there to see if I can conserve a lot of space by doing forever incrementals instead. I wonder if what it is that I'm trying to do is really not financially feasible. I'll read into backup copies further to see how GFS can manage my data that way.

I appreciate the help!

[stamm]

Another comment is that I clicked your link and read through the basics of backup copies - I clearly misunderstood how these work! Now I realize that it's only a delta that gets pushed to the repository that gets merged into the full backup that lives there. That will help a lot!

Also, our big VBK is about 12.5TB and I'll need to contact the Cloud Connect provider to see if there's an alternate way to seed the repository. We have a 1GB Internet pipe but that will still take forever if we're going across the web.

Thanks again!

[veremin]

As far as I'm concerned, many SPs do support seeding into cloud repositories - like shipping a HDD with backup data to them, etc. Thanks.