Add a Backup Copy Job in the SureBackup job

[foggy]

This command allows to add multiple restore points to AG, however using existing AG is not supported.

[TheMonk]

This command allows to add multiple restore points to AG, however using existing AG is not supported.

What would be eth syntax then for adding two or more vms?

If I use Find-VBRViEntity -name "server1", "Server2"|add-VSBViApplicationGroup -name "TESTBRT"

then when I go to create the surebackup job the two servers tell me no backups found?

[veremin]

As Foggy's mentioned, the given commandlet accepts array of restore points, so, you can try get an array of restore points first and then pass it to the commandlet:

Code: Select all

$RestorePoints = @((Get-VBRRestorePoint -Name "Name of first VM" | sort creationtime | select -Last 1), (Get-VBRRestorePoint -Name "Name of second VM" | sort creationtime | select -Last 1))
Add-VSBViApplicationGroup –Name “AG” –RestorePoint $RestorePoints

Thanks.

[TheMonk]

Hmm this don't work as only adds the Second VM

[veremin]

Then, it means that first restore point has not been gotten. Can you just run this code and see the result?

Code: Select all

Get-VBRRestorePoint -Name "Name of first VM" | sort creationtime | select -Last 1

Thanks.

[TheMonk]

I get a result back with that

[veremin]

Can we switch the discussion to the corresponding PS subforum in order not to derail the existing discussion? Kindly, create additional topic describing the problem, and post your code there. Thanks.

[TheMonk]

Can we switch the discussion to the corresponding PS subforum in order not to derail the existing discussion? Kindly, create additional topic describing the problem, and post your code there. Thanks.

My Code?? Im simply trying to use the same code already here, but having issues getting more than one copied VM into a Application group?

I can create different App groups all with different servers using the aforementioned scripts. But the moment I try to add more than one vm it wont do it? I thought this was part of the same thing as this group already?

[veremin]

If we're going to investigate why provided code doesn't work for you, we'd better to that in the proper place in order not to derail the existing discussion with irrelevant PS troubleshooting step, etc.

Thank you for your understanding.

[mdnovak]

When will we be able to run SUREBACKUP for Backup Copy Jobs?

We obviously can do this for regular backup jobs which I do nightly to verify that the backup is good and it will boot properly.

However, I only retain a limited number of primary backups while my backup copy jobs keep backups for a much longer period of time. Also, I use the backups from these BACKUP COPY jobs to copy offsite for long term protection.

Will V9 support SUREBACKUP for Backup Copy Jobs, and if not, when will this feature be added? Without SUREBACKUP, I have to manually test each backup from the BACKUP COPY jobs which is very time consuming and often forgotten.

Thanks

Martin

[Shestakov]

Hello Martin and welcome to the forums!

There's no need to perform recovery verification of copied backups, since backup copy job has built-in integrity check that ensures the copy is bit-identical to the source.

There is a workaround, which is not officially supported though.
Please review the thread for more info and ask additional questions if you have any.
Thanks!

[mdnovak]

The problem in my environment is that on one particular day, my backup copy job failed with:

Code: Select all

6/27/2015 9:10:46 PM :: Error: The process cannot access the file because it is being used by another process.
Failed to open file [\\172.25.75.22\BackupV7X\Backup Copy for DailyC Inside_VM1_Colo\Backup Copy for DailyC Inside_VM1_Colo2015-06-26T203000.vib] in readonly mode.
Failed to restore file from local backup. VFS link: [summary.xml]. Target file: [MemFs://frontend::CDataTransferCommandSet::RestoreText_{4116d911-c2f9-486b-9122-88c48da5f5b2}]. CHMOD mask: [0].
Agent failed to process method {DataTransfer.RestoreText}.

The next day, the backup copy job worked BUT the entire previous chain of backup copy's became unusable. If I could have run a surebackup job against the backup copy job, I would have known right away that there was a problem and not two weeks later when I actually had to restore a file from this backup copy job.

Thanks

Martin

[Shestakov]

Martin,
Have you tried to restore from the source backup file?
I would also suggest contacting technical support to investigate the reason of the error.
By the way, as I mentioned above there is a way to use Surebackup for backup copy.
Thanks!

[haslund]

I understand it makes sense to verify backups before sending them offsite, done!
However once these backups are offsite it would be awesome if we could use them for On-Demand Sandboxes which are ALSO SureBackup jobs.

Maybe we can mess with some scripts, but that is not the right solution.

Please re-consider adding Backup Copy Jobs as the source for SureBacup jobs.

[bg.ranken]

I also would like to have this feature added and would like to present a scenario for why.

Our environment is currently all offsite in our datacenter and we have storage level replication to our DR datacenter. In this scenario our DR has VMware hosts for spinning up server in the event of a full site failover. However until that happens we have VMware hosts literally doing nothing in that site. I would like to use them for SureBackup jobs so that our production hosts are not taxed. We use the same backup storage in both location (old NetApp cluster) so there's no loss of performance for the SureBackup jobs if they were to run on our DR hosts.

I understand why you would want to test your primary backups first before sending them offsite, but since SureBackups check to see if the backup is good there should be no reason you can't use it on the copy since ...

...the copy is bit-identical to the source.

At most your would be one restore point behind if your backup copy jobs are set to run daily, and not even that if you are running SureBackups during the day after the copy is done, which would be doable since you're not hitting any production infrastructure at that point. The administrator would still be notified of the failure and you could design your logic to still perform the same functions it normally would in a failure state against the originating backup jobs.

[crackocain]

Hi

My client have Datadomain 2500. So surebackup jobs deadly slow. Can not finished!

We are moving backups to the NAS via backup copy. But BackupCopies cannot add SureBackup jobs. Why is this restriction? Its not necessary i guess.

Could you add this feature new updates?

[Shestakov]

Hi Mehmet,
As was mentioned above, best practices imply recoverability verification before sending backups offsite.
We will take your request into account though.
Thanks!

[crackocain]

Yeah but Surebackups can't work in Datadomain. Best practises can not relevant this situation.

My client insist to test VM's at Surebackup. How can i worked? I tried to workaround this problem via backup copy. But it can't too.

Production size about 30 TB. I have to use DDboost. Thats why i can not use CIFS repository is primary backup device.

Thanks.

[StephanF]

+1 for Feature Request "Add Restore Points from Backup Copies to SureBackup Jobs / Application Groups"

Explanation:
I fully understand the arguments of Veeam staff that a backup should be verified directly after doing it at the source and not from a copy. And I do understand that backup copy destinations (such as deduplication appliances) can havebad performance.

BUT: Verification of backups is just one use case of SureBackup. Another important function is to fire up a virtual lab for testing something, locking at an old configuration, etc. For us it is often the case that these old restore points only are only available at the long term archive repository.

Please let the user decide if it makes sense to run a SureBackup from a backup copy. You could show a warning that performance may be degraded but do not block the user from doing magical thing with your software.

Thanks
Stephan

[Shestakov]

Thanks for your feedback and requests, they will be taken into account.
By the way there is a possible workaround described in the thread.

[crackocain]

NOTE: repository type cannot be CIFS share.

But secondary backup device is CIFS share NAS.
Only way iSCSI disk mounted to Veeam server but i don't want to.

I suppose thats the only way right?

I think its very critical subject who upgraded backup devices NAS to Datadomain. Especially Enterprise buyers big firms and governments.

[cgranitz]

Please consider adding an enhancement to Veeam B&R that would allow it to run Surebackup jobs against Backup Copy Job data that resides at a remote DR site. Being able to only run Surereplica jobs using replica data minimizes the confidence of the Backup Copy Job data at a remote site. Thank you. - Chris Granitz

[alesovodvojce]

Hi,
we are copying backup data to remote repository server via Backup copy job. This server is then most of the day idle.
Can we use it to run some intensive SureBackup jobs? Or what is the best practice to test the consistency of data in remote copy location?
Regards,
Petr

[Shestakov]

Hi Petr,
Actually best practice is not to use Surebackup for backup copies but verify recoverability prior sending the backups to the secondary repository.
Please read the thread for more information.
Thanks!

[dspjones]

I would also like to request this feature - My use case is a bit different.

I want to be able to use Backup Copy jobs as sources for Surebackup because we use Virtual Labs for Development, Pre-Deployment Testing, Problem/Troubleshooting issues, Testing of patches and other testing scenarios where we want to use like-production systems to test against without affecting the actual production servers. Here's the problem though - We run backups 6 times per day in production. Surebackup for say our exchange vLAB takes about 1:30 to 2 hrs to full startup. Let's say I start up that lab at 11am. Because the lab depends on the backup file it will at 12pm fail and be automatically shut down when the next production backup runs at 12pm. This forces us to do things like temporary backup jobs of the servers in that lab so that they are not locking up the prod. backup files. While this works for me, it is a pain to have to do that AND an unnecessary waste of disk space. It can be a nightmare during periods when developers are asking for test labs for various things left and right. If we were able to use backup copies - I could pick copies from the DR site and have them use those all day long since the hosts, backup files and all the resources there are really not being used most of the day.

The point being SureBackup is not just a tool for testing backups (which you obviously know) - we use it mostly for the lab scenarios so it would be great to give more flexibility in this area with where you can get the servers for the application group like backup copies in addition to what's already offered.

[Shestakov]

Thank you for the feedback, dspjones!
Your feature request is taken into account.

[ekisner]

I would like to put my support behind this feature request as well.

We stagger our backups with different schedules, depending on how important VMs are. There is no significant amount of time during the day where a SureBackup job can run without conflicting with a backup job.

My potential solution was to use cheap storage for backup copies, once per week, allowing a full week of uninterrupted access. I recognize this does not protect from short-term issues, but it does protect from long-term issues, which is more protection than we have now. As I cannot use copy jobs for SB, it isn't an option.

[wa15]

Also adding my request for this feature. I will not go into the details of our case but in some instances, it is not possible to perform a SureBackup of the VMs at source so being able to do this once they are off-site, it would be very useful. Particularly in our global Veeam architecture. Thanks!

[BChavetnoir]

Good Morning all,

Veeam Support - Case # 02311078

We are hosting many applications for Healthcare Domain. We are using DataDomain DD2500 (23 To prod site) to store all backups.
Running SureBackups from DataDomain Repository is a nightware, too slow and are not working at all.

We need to be able to test and check that all applications (Linux and Windows servers) we backup can be successfully recovered by using SureBackup.
Applications we are hosting need many servers and virtual Mabs is required to check if those applications are fully fonctionnal and recoverable.

When I use Backup Copy jobs to a separate server, I'm not able to test and use this source into Applications Groups of SureBackup jobs.
This server has 13 To Repository on a share to store only the Backup Copy jobs.

Could you please help us finding a solution ?

Many Thanks for your feedback.

Bertrand.

[ekisner]

BChavetnoir, while it isn't SureBackup, I would offer a work-around that I presently use.

I wrote a rather interesting script, which iterates through protected virtual machines, always choosing the least-recently checked, mounts the backup to the B&R server, then scans it with our antivirus. Interesting results, solid peace of mind, knowing that I can do a truly offline file scan where an infected system cannot use a variety of avoidance techniques.

Such a methodology could certainly be adapted to run application-specific tests. For example in the case of an SQL server, the script could run a query on the master db of the production server for a list of active databases (and relevant file locations), then iteratively mount them to a test SQL server, run a simple DBCC query. If I remember correctly DNS is just a text file (don't quote me, I haven't looked myself) so similar tests could be accomplished.

Something like AD would be hard to test, so you'd need SB for that, but it's also very easy to isolate your AD backups into separate jobs with different schedules, giving you lots of time to test each.