Veeam B&R OS Upgrade to Windows 2022

[stephenc]

Our two node B&R has been running on Windows 2016 for a while now. In an enterprise wide modernization effort, we need to upgrade these B&R servers to Windows 2022.

What is the best way to do this? Should we build 2 new Windows 2022 VM's and somehow transfer everything over? Or is there a way to re-map the old server's disks to new server?

Any suggestion is much appreciated!

Stephen

[rennerstefan]

Hi Stephen,

please find a guide how to transfer VBR to a new server here:
https://helpcenter.veeam.com/docs/backu ... ml?ver=110

If you also run enterprise manager on that server find the guide here:
https://www.veeam.com/kb1889

You are saying that you would create new VMs so I guess you only run the management roles on those VMs where as proxies and repositories are on different ones!?

Thanks

[stephenc]

Thanks Rennerstefan!

VBR's server is also hosting a proxy and repositories. Would this make the migration process different?

[rennerstefan]

The proxy role itself doesn’t matter really
For repository the question is where that repository stores the data. On a local disk (in your case it would be a vmdk) or on an external devices via Network or San?

[stephenc]

The repositories are additional drives on the same VM, which are vmdk's stored in a SAN.

I guess my question is, how do I "remap" these vmdk files to the new Windows 2022 server I build, and have Veeam recognize them?

Thanks again!

[Gostev]

Just remove these virtual disks from the old VM, and add them to the new Windows 2022 server VM?
This would be a basic vSphere management operations though, so nothing to deal with Veeam at this point.

Then when this is done, just create a new backup repository normally on that new Windows 2022 server VM.

[FrancWest]

You can also do an in-place OS upgrade from 2016 to 2022. In-place upgrades nowadays work fine most of the time. Opinions vary amongst professionals to do it yes or no, but in my experience there’s low risk since the base OS of Server 2016 and Server 2022 is mostly the same.

[stephenc]

That's a good point, I may just try the in-place OS upgrade to save from all the headache. Thanks!

[FrancWest]

Keep in mind to take a snapshot of the entire vm (including the disks where the backups reside) before the upgrade. If something goes wrong you can revert back. When using ReFS this volume will automatically be upgraded to the latest version, which cannot be accessed by Windows 2016 (and Windows 2019) and you end up with a RAW disk when using it on Windows 2016/2019.

[stephenc]

Did not know that about ReFS, thanks for the tip!

[mkretzer]

We also did inplace upgrade of our Veeam Server from 2012R2 -> 2016 -> 2019 and now 2022 - without any issues at all!

[Niksavoy]

Hi there
We also did in place but no luck.
2012R2 -> 2016 -> faled
2012R2 -> 2016 - > some tweeks + VBR config restore - > OK but after 5 days some OS upgrade killed it
2016 clean OS + VBR config restore make the magic

[diegobck]

Hello.

We had upgraded several VBR servers from 2012R2 to 2022 and we had several issues:

- Veeam WMI Name space lost --> VBR Server works properly but Veeam One Server is not able to find any VBR server even if it is the same machine.
- SQL management studio is no longer working --> This is not related to VBR Server behaviour but if you want to take any ddbb bakup you will have to solve it.

For the first issue you have to recreate the WMI namespace (Opened case 05369060), how to solve it:

- Make sure that there are no jobs running;
- Log in to the Veeam Backup & Replication server, open a command prompt as administrator and execute the following commands one by one:

cd 'C:\Program Files\Veeam\Backup and Replication\Backup'
Veeam.Backup.WmiServer.exe /uninstall
Veeam.Backup.WmiServer.exe /install

For the SQL management studio issue you will have to uninstall management tools and then install the last version of SSMS.

Hope that helps.
Regards!

[Andreas Neufert]

Thanks for sharing this here.

[CaptainTightPants]

Hello.

We had upgraded several VBR servers from 2012R2 to 2022 and we had several issues:

- Veeam WMI Name space lost --> VBR Server works properly but Veeam One Server is not able to find any VBR server even if it is the same machine.

For the first issue you have to recreate the WMI namespace (Opened case 05369060), how to solve it:

- Make sure that there are no jobs running;
- Log in to the Veeam Backup & Replication server, open a command prompt as administrator and execute the following commands one by one:

cd 'C:\Program Files\Veeam\Backup and Replication\Backup'
Veeam.Backup.WmiServer.exe /uninstall
Veeam.Backup.WmiServer.exe /install

Hope that helps.
Regards!

Is the WMI Namespace issue presented in Veeam ONE as the "Object properties data collection failure error" message in the Veeam ONE console? Our server has had that message for a little while and I've been meaning to create a ticket but haven't had a chance to yet. If it's the same issue you had then I'll try the WMI commands you used.

[wishr]

Hi David,

Welcome to Veeam R&D Forums.

The Object properties data collection task in Veeam ONE Web Client is not directly related to the WMI namespace of the VBR server. It represents the data flow between the Veeam ONE server (monitoring service, to be precise) and the reporting service. If it's failing, I would recommend opening a support case for troubleshooting. There may be various reasons for this. If you have recently upgraded to V11a, it may be issue #4 from that thread.

Thanks

[CaptainTightPants]

I do believe it started after we upgraded to V11a, and we are also having the high CPU usage mentioned in that thread. I'll create a ticket to get the hotfix.

Thanks!

[diegobck]

Is the WMI Namespace issue presented in Veeam ONE as the "Object properties data collection failure error" message in the Veeam ONE console? Our server has had that message for a little while and I've been meaning to create a ticket but haven't had a chance to yet. If it's the same issue you had then I'll try the WMI commands you used.

Yes, We detected this problem due to we were not able to connect veeam one with veeam server as one was not able to detect any vbr installation. Check first if the name space is created on the wmi properties.

[wishr]

Diego,

When WMI of the VBR server is not accessible for VONE, VONE Thick Client will start raising "internal" alarms and will show the VBR sever as not responding in the Client UI. That's a completely different story compared to what David described above.

Put it simply, in VONE, there are two primary data flows:
1. Between the target infrastructure (VBR, vSphere, etc.) and the VONE server. This way, you are getting most of the data.
2. Between the monitoring service and reporting service of VONE. That way, VeeamONE server passes the monitoring data that is required to build the reporting datasets to the engine of Web Client. That data is used to present the information in the Web Client, reports, web-based dashboards and widgets, REST API, and so on.

An issue with the WMI namespace affects the first data flow. An issue with the Object properties data collection task affects the second data flow.

Thanks