Converting old archive (other save&restore product) to veeam

[JMiG]

Hello,
For a short time we use vSphere and Veeam. The whole system is now on VM, and Veeam works well for saving (full, increment...) on repositery and on tapes.
The next step for us is :
we have a lot of old tapes, written with another software, from a lot of physical servers. We must keep all that data, but we'll soon destroy the old material (tape jukebox, robots, ..)
So the idea is to restore that data to a VM (each server and each period time to time), and save it on tape.
So for each tape (Hera-2005, Zeus-2005, Athena-2005, Saturn-2005, Zeus-2006, ...), I attempt to restore this tape on one small VM (I can load only one disk at a time, ie G:, and the next day I delete the data on tmp-disk, and I restore the data for disk H:, ...). Once the data are on the Disk of temp-VM, I try to use Veeam to create a File-Set on the tape. For me the best is to get a tape (labeled "Archive", or a pair of tapes like "Archives2006-2008") with many file-set "Hera-2005-G", "Hera2005-H", "Zeus2005-F", ....
I try with "file to tape", but it seems it works in incremental mode, it replace the files when I save another year.
I don't find any chapter in the official documentation for this. I'm sure I'm not the only guy to need this.
What will be the best way to do this ?
Thank you in advance for your advices.

[Dima P.]

Hi JMiG,
Unfortunately, Veeam Backup and Replication tape functionality does not support tape media, previously written by any third party software.

[JMiG]

Hello,
Yes it does. Each night we have a save from VM to repository ; each WE the repository is written on tape (a tape library, with robot, input/output slot...).
I try do the same.
I tried first "Files to tape", or "Disk to tapes" - for this a created a job "Disk2Tape". I hope to use this job for all server time to time.
Perhaps I have to create a unique job (and I modify this job to switch between servers). Perhaps I'll have to create one job per server (I prefer to avoid this).
Perhaps I'll have to create a job VM-save to write it on a repository, and then another job to write this repository to tape library...
The result I want : a tape for year 2014 for all servers. If the is 40% full, then I write on the same tape the year 2013 for all the servers... When the tape is full, I take another tape to write 2012, 2011, ...
To manage the tape, I've already created a a special "Media Pool" in the library, named "Annual save" (the media pool internal tapes works well).
J-Michel

[veremin]

I try with "file to tape", but it seems it works in incremental mode, it replace the files when I save another year.

To make it always run a full cycle, just disable incremental processing for it.

The result I want : a tape for year 2014 for all servers. If the is 40% full, then I write on the same tape the year 2013 for all the servers... When the tape is full, I take another tape to write 2012, 2011, ...

In order not to overfill tapes with million individual files it might be worth backing up VMs containing restored data first, and then archive resulting backup file to tape.

Also, it might be a good idea to snapshot the source VMs first, restore data from 2014 to them from tapes written by 3-party solution, backup those VMs and transfer resulting backup files to mediums, using VB&R, then revert snapshot, restore 2013 dataset, backup, archive to tapes, etc.

Thanks.

[JMiG]

Hello,
I tried differents solutions, but some files are locked, or with too long names, or ...
So I've just tried another tool : VeeamZIP. Perhaps it's the good tool : I don't store data nor statistics in the veeam's database, I create a single file in the repository containing the whole VM, perhaps I can rename this file with a significant name like "Save Hera-EndOf2012", and save it to the tape.
But with veeamZIP, first I don't understand the compression : I read "The VM is saved in compressed version", "the empty disk space is not saved", ... In fact, I have an only disk of 1.5Tb, on this disk I have 660Gb of data, and 840Gb empty. I have a repository of 550Gb... The veeamZIP stop at 52% saying "not enough space, disk full". It's curious as compression tool.
Second : veeamZIP can be used only to save entire VM, I don't find any option to save a disk, or a part of disk, like directories. When I create a diretory "Hera2012" and another "Hera2011", I wish to make a save named "Hera2011" and another one "Hera2012".
I'm right on this 2 points ?
Thank you in advance for any idea.
J-Michel