Add a Backup Copy Job in the SureBackup job

[jtupeck]

by Shestakov » Mon Feb 27, 2017 12:45 pm 2 people like this post

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

Has ANY progress been made on this in recent years/versions? Was this really taken into account? I ask, because we have need to run older MS Exchange server restore points inside a vLab so we can pull data off it for discovery using Veeam B&R 9.5 U2, but I am unable to do anything but restore guest files from a Backup Copy Job restore point on this particular server. Sadly, it is an Exchange 2007 box and we cannot use the built in features of Exchange AIR modules and have to use the U-AIR...without the ability to spin up a restore point from a Backup Copy Job, we cant do anything but restore the VM, or VM Guest Files, attach them in the currently running MS Exchange server and then do it all the old fashioned way. Please re-add this request to your list of features that your user base wants, if possible.

[bstilts]

I too am looking for a way to run a SureBackup against a copy job, for no other reason than I do not want the additional overhead of running the SureBackup script in our production environment. It would seem to make more sense to test the backups in our offsite location so that local production resources are not consumed and integrity of the backups at the off site location can be verified simultaneously. Thus, since offsite backups are supposed to be an exact duplicate of the production side, you can be assured that both sides are verified and complete without having to run the additional verification tasks on the production network.

[BChavetnoir]

Good Morning,

Let me introduce myself, we are hosting applications in Heathcare for several hospitals in our region.

Today, I'm still unable to say to my manager that all our hosted applications can be recoverable at yesterday night and be sure
that all applications should work after full recovery. it's critical for us.

We would use Surebackups to be sure that all our applications are recoverable correctly.

Could you please help me finding a solution in creating and managing Surebackups ?

We are using Veeam B&R Ent 9.5 U1 with a DataDomain DD2500 in Repository and creating Surebackups with DataDomain is a nightware.

I managed to duplicate several jobs into a non deduplicated storage for building Surebackups but not so easy to manage.

Now we have a secondary server (srv-bckbuffer) that has 13 To of storage (Windows server)
this server is a W12R2 with SAS drives no Fiber Channel cards for the moment if needed
only two NICs 1 Go to LAN network and another NIC 10 Go directly connected to DataDomain.

From Veeam server I have have a repository to store backup copy jobs in the 13 To of storage (\\srv-bckbuffer\VEEAM) using the 1 Go link.

we woud use backup copy jobs stored in (srv-bckbuffer server) in source of Surebackups jobs but it's not possible for the moment if I have well understood.

which king of repository we should use or configuration to be able to execute Surebackups with lastest backup data (no CIFS repository), is NFS is working and is it the solution ?

Did you managed to use backup copy jobs for source on Surebackups Application Groups ?

for you is it possible to get such configuration working to managing correctly Surebackups in our environment ?

Many Thanks for your help, advices and feedback.

Have a good day.

Bertrand Chavetnoir.

[foggy]

Hi Bertrand, the main issue is that you're using storage devices in the wrong sequence and against our recommended reference architecture. Primary backups should go to a faster raw storage, while dedupe appliance should be used as target for the secondary ones, this is the best practice. This way you would be able to run SureBackup jobs from the faster storage and be sure you're copying recoverable backups.

[jhouse]

We need to use Surebackup to test our backup at a remote DR site. We have a host there and it has all the available resources to run Surebackup on that host. We don't have the resources on the host that has the main backup stored on and could use the resources on the DR site Host. Can this be done with backup copy jobs and Surebackup? Thanks

[Murigar]

I would like to jump in on this also.
I have for the last few years been successfully running my SureBackup at my DR site pointing to Backup Copy data.
I had a dummy job mapped to the Backup Copy data.

This recently stopped working when I created a new backup chain.

I by far prefer to do this testing on idle hardware out at my DR site instead of in my primary VMware stack.

It would be very nice, if from the primary B&R server we could just change the "Source" in the Application Group to match data hosted locally with the Virtual Lab, which happens to be off site.

[Mgamerz]

So... lets say I copy my data to offsite. There is no way for me to verify the backups AFTER the backup copy "verifies the data"? What if I get some bad sectors on a disk or there's some I/O error, I really have no (supported) way to verify my backups at all? Are you serious?

What if there is a disaster at this site and the storage was harmed but still usable - is there really no way to actually test the data integrity anymore? This seems like it should be a key feature - surebackup, making sure your backups are usable - well, okay, only when you put them on your primary storage I guess. You can't test DR.

[Murigar]

Mgamerz

Backup copy data is considered bit consistent with the primary source.
To validate the correctness and protect against bit rot (or bad sectors as you example) the solution is to utilize the "Perform backup files health check..." option in the backup copy job.

The larger logic is that SureBackup will validate the operation of your backups at the primary source and all copies are validated by other methods (such as being compared to the SureBackup validated primary data).

[Mgamerz]

The current reason I am on this thread is because the August 2018 updates messed up our backup server, and I am currently troubleshooting why I could not make a new backup in our forever forward incremental (w/synthetic). Additionally our backup system was pretty fubar'd for almost 2 weeks. I have about 12TB stored offsite from this fileserver, of which I will never be able to replicate due out our rate of change + initial seed size.

However I am still seeing new restorepoints being created on the offsite repository after the new fullbackup was created on the source, but they're from small 10-12GB changes - I would like to be able to boot this VM from our backup copy and perform my own verification of these new restore points to make sure everything is how it should be. But it doesn't seem I can without some significant time investment, I can only rely on data getting there.

This really seems like something that should be a feature, or noted in your documentation as a limitation. Surebackup is advertised as a way to test your VMs for things like updates, but it seems like a pretty huge weakness if I can't actually use my backup copy as a bootable test.

This would have definitely had an impact on if I purchased enterprise edition as the surebackup was one of the big features of that tier - "Hey boss I was able to test our backups at the DR site were bootable".

[Murigar]

I absolutely agree with your frustration and share the same.
May I attempt to offer you a workaround that I use from time to time?

In your described scenario. You could install the VBR console on the remote repository and present local backup data. (Assuming your remote repo is Windows based)
This will make the backup copy data available as restore points on the remote VBR server.

From here you could conduct a restore (or Instant VM recovery) on an isolated network or unplugged nic configuration.
While this will not allow you to do a full file read, you can check operation.

Your other option is to schedule a health check as I noted above to validate bit consistency.
This may take quite some time but is considered best practice to use.

[Mgamerz]

I am using the health check once every other week (as reading 12TB is pretty time consuming). I would do a InstantRecovery, but this server doesn't have 12TB of free space to use as the disks have to be pre-allocated for instant recovery (surebackup does not seem to have this limitation).

I am installing the console on this server, will see if I can follow the workaround, but I am not pleased that I have to do this kind of hackery to get this done.

[witbeckb]

I would also like to see this feature supported. While I understand gostev previously said the source is where verification should happen, there are a number of occasions that make testing at the DR site a requirement. Many of our clients, for example, have resources to run SureBackup at the DR site but not at their primary site. Those clients have the option of surebackup in the DR site (not supported), or no automated testing at all due to lack of resources.

Another reason is auditing. Some of our larger clients need tests done annually on both their DR and source backup chains to pass audits. With our current backup solution (we are still rolling out Veeam) we do those tests manually and were hoping Veeam could automate this. We found the unsupported powershell option from another post, but if it's not built-in and supported by Veeam then we also don't want to rely on it for reporting.

[lightsout]

When I set up an "application group" - I can't pick data from a backup copy job. I'm assuming it is intentional?

So if I wanted to do this, I would need to manually copy the data back to the Veeam server?

Other ideas welcome!

[DGrinev]

Hey,

Testing of backup copies is not the way it's meant to be done.
Please review this thread for additional information. Thanks!

[jmertensnetuk]

I want to setup SureBackup in our offsite location to verify the backup copies.

I have installed Veeam B&R at this location, added the repository, imported the backups and they show as "Disk (Imported)" in B&R console but when I go to add a SureBackup Application Group it does not show the VM's

[nielsengelen]

Hi Jonathan, have a look at this thread for some more information on how this could be achieved via a workaround. We do advise to run surebackup jobs on the primay backup.

[Steven Bricklayer]

Hello,

Any update about that feature?

[veremin]

No updates since Niels' reply. Thanks!

[ASG]

Hi,

Since we are using a secondary destination for all our backups (Backup Copy Job) I would like to run Surebackup from this location rather than the primary destination. Any way to get B&R to do this? Since this secondary Storage gives more speed in IOPS and MB/s (having 72 drives) than the primary RAID inside the Server and the main array is not busy doing the surebackups.

Thanks!
ASG

[Dima P.]

Hello ASG,

It's currently not possible, please take a look at the workarounds discussed in this thread. Cheers!

[ASG]

Hi Dima,

Thanks - but from reading them (I've found it before while searching the forum) I don't think this fits. I'm talking about the same server on the same site with the same instance of B&R - not a Offsite Copy with another B&R Instance. However, I take the "currently not possible" and leave it at that for the moment. Hope that maybe it becomes a possibility in the future since there seems to be some customers that wish for it for years.

Thanks!
ASG

[Cragdoo]

I'd like to add a +1 to the request for this, so the use case is send a BCJ to remote repo, using GFS, and that backup file sits on the remote repo, I would like to, say 5/6/7 months later, check those files have not become corrupt over time. Storage Level corruption guard is meant to address that, but being able to spin up VMs from those "archive" backups, getting the green tick from SureBackup, would be a great confidence boost

[Nick-SAC]

+1 Me Too...

And I’d like to see it as a separately configurable SureBackup Job, so they could be set to run on different/non-concurrent schedules (so as to not require more RAM/resources).

[Dima P.]

Thanks for updating this thread with your votes. Consider all of those being added to this feature request. Cheers!

[Regnor]

If you have different sites and collect the backups via backup copy at the central site, then it would make sense to also run SureBackup at that side.
The configuration would be much easier then setting up a virtual lab for each site and taking care of the routes from/to the masquerade networks.
So +1, Backup Copy should be added as a possible source in an application group.

[bg.ranken]

I want to add the other use case for this is where you have spare hardware and resources to accommodate the virtual lab.

Many people have extra hardware that is unused/underutilized in the DR site ready for a failover. It makes sense to be able to put these resources to use rather than having to use production hardware in the main site or maintain separate dedicated hardware (even if it's older unused hardware it still needs to be maintained).

[johannesk]

Hi,

We use Surebacup alot. The application grouping can be made from a backup job, replication job and a storage snapshot. But is there a way to use a backup that only exists on a copy job to be used in the surebackup job? In our case, the backup jobs have shorter retention time than the copy job for that backup job. Now we need to open a backup that no longer exists on the backup job, only in the copy job.

Is there a way?

Regards,
Jóhannes

[HannesK]

Hello,
I merged your request to the existing thread. PowerShell with instant recovery could be a workaround. Nothing direct in SureBackup.

Best regards,
Hannes

[werner.vergauwen]

Hi

we have a rather dull setup, with 2 DCs. Production running in DC1, non-production in DC2 which also doubles as DR location.
DC1 and DC2 both backup to local repos (SOBR) and then a copy job kicks in to copy the data to the other DC.

Since it makes a lot of sense to us to be able to restore DC1 backups in DC2, we wanted to validate the DC2 copies via SureBackup. Only, we are a little bit at loss as to how this should be configured. The console only allows us to link either a backup job or a replication job, but not a backup copy job. Checking the documentation seems to confirm this.

Question:
* Is this correct? (I sure hope not, but I have little hope)
* Is there another way to make use of the backup copy? Or is this a feature in the making?

thanks
Werner

[HannesK]

Hello,
please use at least forum search if helpcenter is not enough documentation

Best regards,
Hannes