Use old Job backup file as seed for new job

[Zew]

Is this possible?

I want to restructer my backup jobs. Rename them and all. Can I use my old backup .vbk/vib's as seed data. I don't want to have to copy to a portable media, and go all the way to my cold site to re-seed data that's pretty much already there just under a different job/folder.

[foggy]

Yes, you can map the new job to existing backup files.

[Zew]

If I copy the files using a simple direct copy of the files using rsync, there will still be an old .vbm file listed with the name of the old backup job. would I simply delete this file after the transfer and then map the the new job and run it. or keep it there and it will auto rename/remove old vbm files.

In this case I'm planning to move 3 old backup jobs into one job. So there will be 3 different vbm files in the new parent folder.

[foggy]

You can map the new job just to a single backup chain, so VMs from other jobs will be fully backed up during the first job run after mapping.

If you want to get a single combined VBK file containing all the VMs from the three jobs, you can create a backup copy job to merge those three backup jobs into a single VBK and then use that file to seed the new combined backup job.

[Zew]

Foggy your second post contradicts your first one. I don't want to merge the VBK's as they are split out per VM now in v9. So I was expecting to be able to use a .VBK file for particular VMs on the new BCJ as the "seed" and let CBT then copy only changes from that old point in time full backup.

Is this possible? You are stating it will run fulls. I don't want it to run full as it would take forever to get the 200 gigs over my 5 Mbps link. And there is already a sack load of old VBK of the same server already at my cold site storage. So o figured I could refrain from doing the whole create a BCJ to a local portable disc, manually run it over, transfer the data to my new location and re-map the job.

Instead I want to copy .VBK of VM that were in their own folders, move or copy them all into one folder. All these VM would now be placed into one BCJ and I'd hope the job would be able to pick up on the old chains and use them as their "seed".

[foggy]

There's no contradiction, since initially I said that generally you can use your existing files for mapping. However, after getting more details about your scenario, I added that it is not possible to map a single job to several backup chains from different jobs (even with per-VM setting enabled), since there's no combined VBM file for that.

My suggestion was to create a new backup copy job that would have three jobs you want to combine as source and run it locally at remote site (so no need to transfer data to a new location). As a result you will get a new set of VBK files (per-VM) for all VMs in those three jobs and will be able to map your new remote combined job to it.

[Zew]

Thanks for all your help so far on this foggy. However when you say create a new BJC with my original BCJ as the source. Does this mean I'd need to run a copy of veeam at my cold site? Right now its simply proxied through a Debian server (ssh) which is NFS connected to a storage unit there. This proxy wouldn't be able to complete this task eh?

[foggy]

You do not need a second instance of Veeam B&R at the cold site, you can create the new job on your main Veeam B&R server. Please note, that you cannot specify a backup copy job as a source for another backup copy job directly, you would need to specify the corresponding backup repository instead.

[Zew]

but then I'd be back to the old school style. Build BCJ pointing to original source, local repo. Stop the BCJ, move the data to a mobile device (generall a USB HDD) manually move it to the Cold site, Copy the contents to my NAS. Adjust permissions. Scan the repo, and then map the job.... basically defeating teh exact purpose of my initial question. use old backup data for "seed" of new BCJ

[Zew]

There's no contradiction, since initially I said that generally you can use your existing files for mapping. However, after getting more details about your scenario, I added that it is not possible to map a single job to several backup chains from different jobs (even with per-VM setting enabled), since there's no combined VBM file for that.

My suggestion was to create a new backup copy job that would have three jobs you want to combine as source and run it locally at remote site (so no need to transfer data to a new location). As a result you will get a new set of VBK files (per-VM) for all VMs in those three jobs and will be able to map your new remote combined job to it.

It be really nice (I find cataloguing to be a lacking feature in Veeam) if you could copy (or move) and then point a chain of (Per-VM) Chains in a folder. and then have Veeam generate a VBM for a new job. Thus keeping your entire old chain and retention points while allowing you to clean up job names, or concatenate together what used to be separate jobs. I originally had seperate jobs mainly cause at teh time you guys didn't provide per-vm vbk files. Which now that you do i feel I can clean up my jobs better. however cleaning up jobs with linked BCJ seems to be rather painful task to complete.

[foggy]

but then I'd be back to the old school style. Build BCJ pointing to original source, local repo.

Not local, but remote. You need a backup copy that will run from remote repository to another repository located at the same remote site. And then no need to:

Stop the BCJ, move the data to a mobile device (generall a USB HDD) manually move it to the Cold site, Copy the contents to my NAS. Adjust permissions. Scan the repo, and then map the job.... basically defeating teh exact purpose of my initial question. use old backup data for "seed" of new BCJ

[Zew]

Hey Foggy.

So I again did a bunch of research and my minds been tinkering away. And sure enough with enough googling I found out pretty much what I already knew the problem was. Hoever when you told mme its not possible. It is but I'm sure it's not supported?

powershell-f26/creating-vbm-from-scratch-t20568.html

tsightler posts

"Creating a VBM can sometimes be useful. One common use for me is to create a seed for backup copy jobs from a "normal" backup chain. Normally a backup copy job must be seeded from a single restore points/VBK file which prevents you from using an existing backup chain for a seed since these almost always have multiple chains, however, using this script you can grab just the most recent VBK from an existing backup chain, import it, run the script to create a VBM, and then combine them to use as a seed for a BCJ."

Thats exactly what I need! and it be relly nice if Veeam implemented this in the GUI or something. Anyway. I already started attempting this a more manual and supported way by first. Taking my old backup job that had my PDC and renaming it to what I want me new Job to be in this Case Active Directory. Following this.. veeam-backup-replication-f2/rename-back ... 17216.html

I then started to follow your advice from this old post... veeam-backup-replication-f2/question-ab ... 13861.html
So I started to add my secondary DC to the newly named job. And this is again where the issue starts.. I can copy the old chain to the new folder location of the existing job, but as we all know the VBM file in this new job has no idea about this existing chain. Thus adding this VM to this job I can't "Map" the job since it want s to Map to a VBM not a folder per say. So if I picked the VBM of the job that holds this VMs chain, my other chain holding the PDC chain will be lost.

So I'm very close to getting this done. I just need to see if I can indeed manage to create my own VBM to complete what I want to do. Or suck it up and have the additional 2 DC's simply have their own new chains created....

[foggy]

Yes, operations with VBM files are not supported, however no one prevents you from trying. That said, the suggested above way of combining the jobs with the help of a temporary backup copy job looks easier to me.

[Zew]

Be nice as a feature request to be able to easily rename, and concatenate jobs. specially now that chains are stored per VM. being able to "clean up" either your own or someone elses mess easily would be a great feature. however, I guess this could also provide jobs for the technically abled..

[Zew]

I attempted to follow warpconduit's suggestion with a couple powershell commands

"$BackupName = "OriginalName"
$Backup = Get-VBRBackup -Name ($BackupName + "_imported")
$Job = Get-VBRJob -Name "$BackupName"
$Backup.Update($Job)"

what I found was i was getting a null value when I specified the variable object attempting to select the job name with + _imported. So my backup variable I specified it with just the job name and it returned a proper value. I then followed the rest but changed the last part (calling the objects update method) to my new job VS my old job name. sure enough I hit f5 and it removed from imported and showed under disk. However there were now 2 instances of my new Job name, one holding teh chain of my PDC and 1 instance of my other DC, and the other holding my old chain (of my secondary DC). Man getting chains together has always been a PAIN. PLEASE PLEASE PLEASE VEEAM bring in a option in the gui to combine jobs or at least use old chains for new jobs (I know this is possible if you have an existing VBM, but not if you want to merge an old chain that isn't in the new or renamed jobs VBM)! or re-named jobs!

[ober72]

Sorry might have missed this, but what if I want to change a job to a per-vm job. If I create a new job in a repository that is per-vm then map the new job to the old job will it then in the new job be a per-vm vbk job.

I am having issues with a very long merge after a job.

thanks

[ober72]

one other question if you seed from a job that was created in Veeam 8 with no bitlooker will the new job use the bitlooker for the whole vm? My understanding is that If you turn on the bitlooker on a job that was created in Veeam 8 that was just upgraded it would need to do a new full backup to avoid the dirty bits.

[foggy]

Sorry might have missed this, but what if I want to change a job to a per-vm job. If I create a new job in a repository that is per-vm then map the new job to the old job will it then in the new job be a per-vm vbk job.

Do you mean copying the old single full backup chain to the repository configured for per-VM chains and then mapping the new job to it? It will continue as not per-VM until the next active full.

one other question if you seed from a job that was created in Veeam 8 with no bitlooker will the new job use the bitlooker for the whole vm? My understanding is that If you turn on the bitlooker on a job that was created in Veeam 8 that was just upgraded it would need to do a new full backup to avoid the dirty bits.

For BitLooker it doesn't matter what do you have in existing backup files. When you enable BitLooker on a job, it will do its job during the next cycles, however, only for data copied within the particular job run. So if you want it to process the entire VM, you need to perform an active full.

[ober72]

Thanks Alexander that answers both my questions.

[nrexpress]

You can map the new job just to a single backup chain, so VMs from other jobs will be fully backed up during the first job run after mapping.

If you want to get a single combined VBK file containing all the VMs from the three jobs, you can create a backup copy job to merge those three backup jobs into a single VBK and then use that file to seed the new combined backup job.

I know this is an older message, but I am wanting to do what is in this message -- I want to get a single VBK file containing multiple VM's to see a new combined backup job. How do I do that? Whenever I use the "map backup" (on the storage tab of the backup job), it only allows me to select only from existing backups, not from existing backup copy jobs.

[PTide]

Hi,

This is by design. It is not possible to map a backup job to a backup copy job chain since U3.

Thanks