[VHRISO] feature request for replication

[Zedification]

MOD: split from veeam-backup-replication-f2/managed-har ... 96192.html

Hi all,

As a feature request, for VHR, I don't know if this is the right thread, and if we could directly from the VBR console control the backup replication from one to one or more VHRs.
A bit like Data Domain or StoreOnce replications.
I've already done it with rsync. It works well. I tested a failover scenario, and mapped the backups to the second site, and did a rollback to the main site.
What do you think?

See you ..,
Zed!

[Gostev]

Just use Backup Copy jobs. rsync is a bad idea as it has no awareness of VBR activities which might be still processing and updating the backup files, it really struggles with large file etc. while not bringing any meaningful benefits.

[Zedification]

Thanks for your replay Gostev.

OK to use Backup Copy, but I don't find the way to map the Backup Copy chain to a backup job.

I want to continue to backup with the same job when I do the failover to the recovery site.
Is they're a way with a backup copy job?
Thanks

[HannesK]

Hello,
backup jobs cannot be mapped to backup copy jobs to continue a backup chain. That is a open feature request for scenarios where the backup target is not available anymore and customers want to continue on the backup copy target without having enough space to start a new backup chain. I add you as +1

besides the things Anton mentioned above:
1) how much data do you copy with rsync?
2) do you use fastclone / reflink on XFS with GFS restore points and how fast is that?
3) as far as I see, rsync does not support reflink / fastclone. Is that wasted space not a problem for you`?

block-based replication (e.g. DRBD) would solve 3 and probably the other concerns. But there is currently no plan to add such a replication in Hardened Repository ISO. Getting an option to map Backup Copy Job to Backup Job is more likely to be implemented.

Best regards
Hannes

[Zedification]

Thanks for your reply Hannes,

1) how much data do you copy with rsync?
We are talking about several dozen TBs.
It's a fresh install include the VMware part, physical machines, etc.
So I don't have yet the final size of the backups

2) do you use fastclone / reflink on XFS with GFS restore points and how fast is that?
Yes I do. For now I'm just testing a few backup chains, which aren't very large. I didn't really have a problem with copy time.

3) as far as I see, rsync does not support reflink / fastclone. Is that wasted space not a problem for you`?
It will be a problem, because the sizing of both VBR was calculed with the fact that the fasclone gain space.
If that don't work, I have find an other way. What is DRBD?
What is the road map to add Backup Copy Job to Backup Job? V12.3? hope so !

Regards,
Lucien

[HannesK]

for DRBD, I would like to refer you to your preferred internet search engine. It's out of scope from Veeam side. I just expect it to work because it's block-based and I remember that I used it 20 years ago in university...

Mapping BCJ to BJ has no timeline for now. 100% not in 12.3.

[Zedification]

OK, I will search for DRDB.

Thanks for your answers