Feature Request: tape multistreaming / per disk backups

[vmtech123]

For large VM's, (40+TB) I have a bit of an issue.

I am backing up to repo, doing copy jobs, and also have Tape at a few sites.

LTO8 and Veeam is getting me FANTASTIC speeds. but, 40TB is still 40TB.

Now, the only way I know to split this would be to create 2, or maybe even 3 backup jobs for the one VM. The only issue with this is that I can't restore to original location. There are obviously some HUGE files in here, and I don't always want to be restoring to another location and copying it to the original. Due to file sensitivity, permissions, network rules etc this can become tricky.

Would backing up the system disk with each job solve this?

If the whole VM goes kaput, can I pick each job and click "restore virtual disk" and get it back?

Is there a way perhaps in version 12, you have an option to "Split the backup job in to 2" or something to split up the VMDK's into jobs? Similar to doing per VM jobs, but with PER VMDK. Allowing us to pick what VMDK's are in which job would be ideal as I could group a few smaller disks together and keep the huge ones separate.

My main reason would be putting this to TAPE as once it is all in a single file, it ends up taking longer than 24 hours and the jobs fail. If I say Prevent from being interrupted, things really fall behind and that isn't great either.

I feel like this would be relatively easy for you guys and an amazing feature to have large VM's backup faster, use more streams, and get on to tape utilizing multiple drives at once.

It would also solve the issue of having to manually select the drives to backup, and potentially miss one as it would be excluded after I add another.

Thanks

[HannesK]

Hello,
having parallel streaming for tape or having per-disk-chains for the reasons you mention are existing feature requests and we add your vote to these requests. These feature requests are more complex than it might look

With "restore to original location" you mean VM restore directly to the virtual infrastructure? Just for curiosity: as you have disk backups and BCJ... how often do you restore old backups from tape directly into the production infrastructure?

Best regards,
Hannes

PS: I moved your request to the tape forums and edited the title to make it easier to read

[Dima P.]

LTO8 and Veeam is getting me FANTASTIC speeds. but, 40TB is still 40TB.

That's a vm or physical host? Can you share what type of data you store?

Would backing up the system disk with each job solve this?

It depends: if you need to backup of specific huge files maybe you can use file to tape job and tape those files out directly from the source machine.

If the whole VM goes kaput, can I pick each job and click "restore virtual disk" and get it back?

Should be possible to restore single disk from the vm backup (with or without other disks being excluded).

Is there a way perhaps in version 12, you have an option to "Split the backup job in to 2" or something to split up the VMDK's into jobs? Similar to doing per VM jobs, but with PER VMDK. Allowing us to pick what VMDK's are in which job would be ideal as I could group a few smaller disks together and keep the huge ones separate.

As Hannes already mentioned, we will add your post to this feature request!

[vmtech123]

Hi Dima and Hannes.

Thanks for the replies.

VM's are what I am backing up. Mostly large file servers with many disks. As we have B&R Enterprise, paying for additional licenses to do file server backups / NAS backup isn't really in the question right now. We will go that route if we want file versions kept. (something I'd like to implement in the future). Time based backups are good for now, however it makes restores "interesting" sometimes.

[Dima P.]

SP,

Thanks for the details! As Hannes already mentioned we will add your vote to per-disk processing request.