questions about backup copy job

[davids355]

I have just been experimenting with a backup copy job.
For the main backup the VBK file is around 156MB and the incremental vib files are around 30MB.

After running the veeam copy job for the first time it has produced a vbk file of 30MB and after a second sync, produced a vib file of around 40MB.

Just wondered why the vbk is so much smaller for the backup copy job. Does it perform higher amount of compression?

And how would files be restored from the copy job, manually copied?

Lastly, is a copy job suitable for running over WAN? And what sort of speeds would be required?

In production environment I have around 7 VMs that average around 300GB each. I am backing these up to a local repository and I would like to set up a backup copy job to copy them to remote repository which will be running over a 10MB link. Would that be feasible?

Thanks
Dave

[Shestakov]

Hello Dave,
Testing with small files is not that demonstrative in terms of size changes. But what you experienced could be caused by great deduplication in case of Full backup and changes in data blocks for .vib. Anyways 10MB is not a big difference.
I would suggest trying to backup real VMs.

And how would files be restored from the copy job, manually copied?

You can perform restore from backups and files of backup copy job just like restore of basic backup job. Is that what you were asking about?

Lastly, is a copy job suitable for running over WAN? And what sort of speeds would be required?

Yes, copy job along with replication job can leverage WAN acceleration.
There is no requirements for bandwidth, on the contrary, WAN acceleration shows better results for slower network speed.

In production environment I have around 7 VMs that average around 300GB each. I am backing these up to a local repository and I would like to set up a backup copy job to copy them to remote repository which will be running over a 10MB link. Would that be feasible?

Absolutely. And yes, it makes sense to use WAN accelerator using 10Mbps link. Thanks!

[davids355]

Hi,
thanks for the quick reply.
It is a real VM but its just a small linux box running pfsense.
I take your point though, now that I understand it a little more I will run a larger windows VM and see how it runs.

Thanks you have answered all of my questions!

[Shestakov]

You are welcome David.
Try out production VMs and if you have questions, don`t hesitate asking!

[davids355]

Hi,
I do have another question related to this.

I have a backup job running that backs up hyperV VMs to an offsite repository.

I want to change this so that it backs up to a local disk.

So I understand that I can copy/move the backups to that new (local disk) location then change the settings of the job to use the new (local) repository.

I then want to set up a copy job for each VM so that it copies the backup to the remote repository. I figured that was a more sensible way of setting it up so that I have fast local backups and off site copy for redundancy. Am I right in thinking that is a good way of doing things?

What I also want to know is, seeing as I will have a fully up to date copy of the backups on the remote repository, can I make use of them to "kick start" the copy jobs or do I need to allow the copy job to take the entire backup off site from scratch when I set this up?

[Shestakov]

I have a backup job running that backs up hyperV VMs to an offsite repository.
I want to change this so that it backs up to a local disk.
So I understand that I can copy/move the backups to that new (local disk) location then change the settings of the job to use the new (local) repository.

After you move existing backups to a new location, you need to rescan the repository and map the job to the backup files.

I figured that was a more sensible way of setting it up so that I have fast local backups and off site copy for redundancy. Am I right in thinking that is a good way of doing things?

Absolutely. That`s what we recommend to do.

What I also want to know is, seeing as I will have a fully up to date copy of the backups on the remote repository, can I make use of them to "kick start" the copy jobs or do I need to allow the copy job to take the entire backup off site from scratch when I set this up?

Didn`t catch the question, could you clarify your goal please?
Thanks!

[davids355]

I am going to set up a backup job that backs up VMs to local harddrive on the host server.
Then I am going to set up a copy backup job that will copy those backups to the remote repository.

The problem for me is that the initial backup copy job is likely to take quite some time - I think there is around 600GB or so of data to back up and I can only get around 10MB/s to the off site repository.

However, I have not made these changes yet and as it currently stands I am simply doing a standard backup job direct to the off site repository, so it already contains a backup copy of all the VMs.
As such I wonder if I can utilize those existing copies of the VMs when I set up the new backup copy job rather than having to run the backup copy job from scratch.?

[Shestakov]

Got it. Thanks for the explanation.

Yes, it`s is also possible, the process is called Backup Seeding. Please take a look at the description.
Thanks!