Problem with the AIR Functionality

[Maieu_san]

Hi All,

I'm having some trouble at a customer site.
I'll try to explain the situation..

Our customer has 2 sites. Site A and Site B.
Site A:
- vCenter with 3 ESX hosts.
- All production VMs
- 1 Veeam Proxy server

Site B (DR site):
- vCenter with 3 ESX hosts.
- Replica VMs
- Physical Veeam server

Connection between Site A and Site B is Active – Passive.

The original idea was to take backups over a dedicated line from Site A to Site B.
Because the line between the 2 sites is not fast enough, it was impossible to do these backups from Site A to B.
Because of this issue we became creative. We do replication between Site A and Site B, and take Backups of the Replicated VMs.
This all worked pretty well, until we tried to use the AIR functionality (U-AIR, AD-AIR, SQL-AIR).

Since we do not have backups directly of the running production VMs, it seems impossible to use this.
My guess is the following:
When using the AIR wizard, you need to specify the IP address of the running production machine for which you would like to restore an item.
Veeam then searches for the backup file for this machine.
When the backup file is found and the vLab approved, you can restore your item.

In our case, we don’t have a backup file of the running production VM, but of the offline VM Replica!
I assume this is the reason why I am unable to get this to work. But can someone confirm this?
Or better yet, is there a workaround for this issue? Will we be able to use the AIR Functionality is our setup?

Thanks in advance!
Best regards,
Steven

[foggy]

Because the line between the 2 sites is not fast enough, it was impossible to do these backups from Site A to B.
Because of this issue we became creative. We do replication between Site A and Site B, and take Backups of the Replicated VMs.

Actually, the size of the transferred data should be the same for backup and replication jobs in case of similar settings and architecture. Just make sure you have proxy servers deployed on both sides to ensure the compressed traffic is sent over the link.

Since we do not have backups directly of the running production VMs, it seems impossible to use this.

What kind of errors do you get?
I guess it should work if you add VMs from backups rather than from infrastructure when creating application group.

[Maieu_san]

Foggy,

Thanks for your reply, and sorry for the late response on my side.

I Add VMs from backups rather than from infrastructure.
But the problem is that these backups are backups of the replica VMs.

I'll try and give an specific example.
In Site A is a production VM : VM_AD
In Site B is a replica of this VM : VM_AD_Replica
The Veeam Server is located at site B.
Since we do Replication from Site A to Site B, and take backups of the Replica's. I have a backup file called VM_AD_Replica.vbk

I have created a SureBackup job to test the backup, with as Application Group the VM_AD_Replica.vbk backup as backup file.
When I run the SureBackup job alone, this works perfect, and the job is successfull.

Then I want to test the Active Directory Application Item Recovery.
When starting the wizard, you need to give the Domain Controller IP Address, which I provide (This is the IP Address of the VM in production on Site A), I give a username/password and click Connect.
The Resolved DC comes up.
Next in the wizard, you need to specify a Restore Point. Which is something I actually cannot provide since we do not take backups of the VM_AD, but of the VM_AD_Replica!
But for finishing the Wizard, I choose latetst restore point.

The Lab Request is now visible in the Enterprise Manager.
When Approving the Lab Request, you need to go through another wizard.
The 3th step in the wizard, I think, is 'select backup to use if VM is processed by more than one job'.
But since we do not have backups of the production VM directly, what do I choose here?!

In the SB job I am using the VM_AD_Replica.vbk, but during the Active Directory Application Item Recovery I need to specify another Backup file? On of the Production VM, which we do not have..

Does this make sence to you?
I tried to explain it as clear as possible, but it isn't easy

Hope you can help!
best regards,
steven

[foggy]

Next in the wizard, you need to specify a Restore Point. Which is something I actually cannot provide since we do not take backups of the VM_AD, but of the VM_AD_Replica!

Valid point. I need to check how the wizard detects the required backup to use, not sure whether it is able to locate the replica VM backup automatically.

The 3th step in the wizard, I think, is 'select backup to use if VM is processed by more than one job'.
But since we do not have backups of the production VM directly, what do I choose here?!

Do you see the same replica backup file (VM_AD_Replica.vbk) in the list? If you could point the wizard to it, I assume this would work.

[Maieu_san]

Foggy,

The problem is that I do not see the VM_AD_Replica.vbk in the list.
And I assume this is the reason the AD-AIR does not work..

[foggy]

Yes, you need somehow let it "know" what backup to take... And currently I do not see how this could be done in the scenario with virtual lab requests.

However, you may try the simplified workflow (described in U-AIR guide, p.8 ). If you open the running SureBackup job session, right-click the VM, and select Active Directory item recovery, you will be able to specify IP addresses for both backed up VM and production VM. Probably, this will work for your case. I would appreciate if you try that and update the thread with the results.

[Maieu_san]

Foggy,

I tried to follow the user guide, but when I right click the VM in the SureBackup job all the options are greyed out.

The SureBack job has run successfully before, so I do not think there is a problem with it.
I'll run you through how I set it up. Maybe I missed something?

I created an application group with a VM added from backup file, which is VM_AD_Replica.
I gave it DNS Server, Domain Controller and Global Catalog as Role.
Then I created a SureBackup job with the Application Group and No Backup job. Since the Backup file is in the Application group I thought of this as unnecessary?
When I run this, it finishes successfully.

But when I try to follow the Guide, I cannot choose the options because they are all greyed out.

Am I doing something wrong?

Best Regards,
Steven

[foggy]

You need to keep the application group running in order to be able to perform AIR. This is done by enabling the corresponding check box in the SureBackup job properties, at the Application Group step.

[Maieu_san]

Foggy,

I can open the Active Directory Application Item Recovery the way you said.
But... I still have an issue.

I'll try to explain.
I run the SB Job, wait till it is finished and keep the application group running.
Then I go to the Active Directory Application Item Recovery by right clicking in the SB Job.
Till here, everything goes well.

When the AD-AIR Wizard opens, I need to give 2 IP addresses.
1 for the DC in the vLab, and 1 for the Production DC. I think, since the DC in vLab and the DC we want to restore to is the same, I need to give the same IP twice?
Is this correct?

I ask because I have something strange happening.

It seems that when I click next and go to 'AD Object', I go to the production DC?
I think this because I asked our customer to create a testuser. I ran a Replication job and a Backup job so the user would be in the backup file.
Then deleted the user, and ran the SB job with the latest backup file.
And now I have a problem... In the 'AD Object' I do not see the user!
When I log in to the console of the running SureBackup VM, and go check the users in the AD the Testuser is there!

So, I know the user is in the backup but it does not show when I use AD-AIR. When I log in via the Console in ESX, I see the user as expected.

I think I'm doing something wrong, but I don't know what it is.
Do you have an idea what could be wrong?

thanks!
best regards,
Steven

[foggy]

Steven, I guess you need to specify masquerade IP address for DC in the virtual lab.

[Maieu_san]

Foggy,

I tried that, but I get the following message:
The server is not operational.

But it is running, I can connect via console, I can ping it from the Veeam server, ...

What am I missing?

[foggy]

Well, I suggest asking support for assistance as I had never tried to reproduce such scenario myself.

[Maieu_san]

I logged a call at support for my issue.

thanks for your assistance!

[foggy]

Would appreciate it much if you could post back about the results. Thanks!

[rbrambley]

It sounds like you have a slight mix up and are very close to getting it to work. Maybe watching this AD U-AIR lesson from our Online University will help you see your error?

http://www.veeam.com/university/uairad.swf

[Maieu_san]

Rich,

Thank you for your reply, but I have tried to run the AD-AIR wizard as explained in the lesson.
As described above this does not work for me. I am, however, able to run the AD-AIR wizard directly from the SB Job (Thanks to Foggy's explanation).
But even then it does not work for me, since I seem to be unable to use the correct backup file to restore.

As Foggy suggested, I logged a call at support for which I uploaded the logs today.
I'll keep you posted if I get some new info.

[Maieu_san]

Foggy,

you asked me to keep you up to date in regards with this problem.
I'm just of the phone with Veeam support, and it seems we have found what I did wrong.

It was a mistake on my part. I thought the Masquerade IP, was the IP of the VM in the vLab that you can see in the VMware vSphere Client.
I now know that you need to see in your vLab for the Masquerade IP Range, f.e. 10.255.10.D
And you need to take the last digit of the IP of the VM in the vLab, f.e. 10.10.10.100
The Masqueraded IP address is 10.255.10.100, where I would use 10.10.10.100.

When I used the correct IP address I was able to do the restore.

I would like to thank you and Veeam Support for the fast responses and the good help!

best regards,
Steven

[foggy]

Glad to hear you've finally nailed it, Steven! Thanks for posting back, now we know that the scenario is possible to implement.