Veeam Backup & Replication 12.3 just released

[m.novelli]

Generally I receive an email from Veeam with security notifications, this time didn't happen yet

Marco

[Gostev]

Those notifications you're thinking about have always been for critical vulnerabilities only (i.e. when you need to hurry to upgrade, just as you said).

You need to be subscribed to Security Advisories to get security notifications irrespective of their severity.

[robnicholsonmalt]

Hello Rob,

We will review the existing free space notification with the team to make sure it provides accurate estimates. Thank you for the detailed investigation!

Thanks - can appreciate things like this are very tricky to calculate. It did prompt me to look at Gparted which is much easier to use than I expected. I typically commission servers with 100GB OS drives. Might up that a bit in future.

[m.novelli]

Those notifications you're thinking about have always been for critical vulnerabilities only (i.e. when you need to hurry to upgrade, just as you said).

You need to be subscribed to Security Advisories to get security notifications irrespective of their severity.

I've subscribed, thanks!!

[t.gebler]

I've just done a test install and upgrade on a virtual machine. Free space started off at 30.04GB and I got the warning. I knocked up a bit of PowerShell to monitor free disk space during upgrade and it dropped to 26.42GB minimum and then finished at 28.46GB. So I suspect the warning is based upon a full install, not an upgrade. My upgrade needed 4GB free disk space during the upgrade and then used an additional 1.5GB for the new version.

This is a real pain for us currently, specially in Cloud / IaaS environments, disks tend to be small because space is paid per minute. Is there a way to skip this free disk space check?

[robnicholsonmalt]

It's only a warning, it can be ignored.

[BackupBytesTim]

So I do want to take a moment to comment that I definitely approve of the migration to PostgreSQL. The only particular disapprovement of the migration that I would comment on is that installing it in an environment where it's not used yet just bloats the software even more. It still perplexes me how Veeam's software takes up more space than Windows itself does, like how does backing up anything require more software than the entirety of Windows to do all its things, but unfortunately that's how it is, so I've accepted it, and I continue to nag people for creating Windows VMs with 40 GB of disk space these days, really when was the last time 40 GB was "best practice"?

Overall I think the move to PostgreSQL is an improvement, and I agree with @Gostev on eventually wanting to have a more "appliance like" feel to it. It should more "install and done" without all the maintenance of individual components like a VBR server here, proxies there, repositories in a few locations, and a database on that spare server in the main office. Definitely like the fully self-contained installation idea.

As a suggestion along that route, and perhaps this is the end goal:

In my opinion a fully configured ISO that can be used a VM's virtual disk would be a good solution here, have everything already installed and configured, more or less, so we just drop the ISO into whatever virtual environment we have, whether that's cloud service or on-premises, boot it up, web interface for management that will prompt for initial configuration on first boot. That's how Acronis does their self-hosted deployment, noting they push very strongly towards their cloud-hosted deployment, but still support and develop self-hosted options for most workloads.

[robnicholsonmalt]

It still perplexes me how Veeam's software takes up more space than Windows itself does

I've never really thought about but, yes, you're right - it does seem a little OTT. Then again, Backup Exec was a pretty big install as well.

[scott.anderson]

I've just done a test install and upgrade on a virtual machine. Free space started off at 30.04GB and I got the warning. I knocked up a bit of PowerShell to monitor free disk space during upgrade and it dropped to 26.42GB minimum and then finished at 28.46GB. So I suspect the warning is based upon a full install, not an upgrade. My upgrade needed 4GB free disk space during the upgrade and then used an additional 1.5GB for the new version.

But CAVEAT - this is a very minimal test and you take your own chances in ignoring the warning....

Veeam install includes all of the plugins for Azure, AWS, Google Cloud, Nutanix etc. If you don't actually have any use for all those plugins you can uninstall them to get space back on your C:
This has been a growing "feature" of the VBR install for a few years. What would be really nice is if you could unselect them from the install (you can always add them back in later if you have a new requirement for them) or select a different location for all these installs that currently have to sit using space on the C:.
They are all a few Gb in size, so by the time you have added them all up you can get as much as 15Gb of space back on your C:, which can be the difference between getting the 43Gb free space for the upgrade and not getting it!

[m.novelli]

Definitely we need an option to unselect many features... to free space and speedup the installation, it's soooooo slow

Even better for most environment Internet connected... a small installer that just download the feature you select during setup

13 GB ISO is quite unfair...

Marco

[mjr.epicfail]

I think a lot is interwoven and cannot be uninstalled without breaking some stuff.
But it would be nice to get a smaller footprint.

[ajhskao]

Just a quick check to see if anyone else has seen this. Upgraded from 12.2 to 12.3, we just have a single central backup server (Windows Server 2016). After the (apparently successful) upgrade, my jobs using SOBRs fail, saying "Unable to allocate processing resources. Error: No scale-out repository extents are available". I'm in the process of rebooting (which has triggered a queued Windows Update as well, duh), to see if that helps. Of course this is taking ages......zzzzzzz
Thanks all.

[Gostev]

At the moment 12.3 does have any known common support issues that are a result of regression in the existing functionality (thus there's no usual Top Issues sticky topic still).

[brent-h]

My upgrade from 12.2 to 12.3 last week worked (I saw the forum posts and temporarily re-enabled Windows Script Host before I first ran the installer). For future versions, the use of Windows Script Host (in the PostgreSQL installer) will become more of an issue, since Microsoft is deprecating (and later removing) VBScript ( https://techcommunity.microsoft.com/blo ... ps/4148301 ). The people that make the PostgreSQL Windows installer that Veeam appears to use seem to already be aware of this ( https://github.com/EnterpriseDB/edb-ins ... /issues/98 ).

Process logs seem to show just one use of cscript (WSH):

Code: Select all

cscript.exe //NoLogo "C:\Program Files\PostgreSQL\15/installer/server/initcluster.vbs" "NT AUTHORITY\NetworkService" postgres postgres C:\Users\usernameredacted\AppData\Local\Temp/postgresql_installer_e551ec2be3 "C:\Program Files\PostgreSQL\15" "C:\Program Files\PostgreSQL\15\data" 5432 DEFAULT 0

Perhaps someone from Veeam could assist PostgreSQL/enterprisedb with porting initcluster.vbs to PowerShell, thereby eliminating the need for windows script host in later updates to PostgreSQL/Veeam?

[tommyw]

Hello,
Can someone tell us if Veeam supports SAN Clients and Fibre Transport? Something like Veritas SAN FT Client features for Windows and Linux client servers?

Thanks.

[Andreas Neufert]

Yes, a similar feature is available for Windows Agent and VMware backup workloads.
In Veeam this is called Backup from Storage Snapshot.
https://helpcenter.veeam.com/docs/backu ... ml?ver=120
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

[ajhskao]

Just a quick check to see if anyone else has seen this. Upgraded from 12.2 to 12.3, we just have a single central backup server (Windows Server 2016). After the (apparently successful) upgrade, my jobs using SOBRs fail, saying "Unable to allocate processing resources. Error: No scale-out repository extents are available". I'm in the process of rebooting (which has triggered a queued Windows Update as well, duh), to see if that helps. Of course this is taking ages......zzzzzzz
Thanks all.

It took me two reboots before things were working again. After the first reboot, jobs would run, if started by hand, but only after the second reboot would any scheduled jobs run, or jobs chained to other jobs. Very odd. I'm putting it down to the server being in the middle of Windows Updates for now.

[chrisr]

The other point to forcing everyone to use postgres is one of familiarity and support-ability - not every dba knows anything about it so will require training, and I've even had issues trying to get other opensource software past the cybersec guards in the past. I'd agree that having the ability to untick all the unused features/explorers etc would be a very good thing to have

[robnicholsonmalt]

Not sure this is related to the 12.3 upgrade (agent v6.3.0.177) but just restored my workstation back to Friday night backup because I messed up Windows. It sat for over 1.5 hours attempting to restore the recovery partition. I tried cancelling it but it failed to cancel (waited 15 minutes). Restarted the workstation and it booted fine. So this means that either is didn't touch the recovery partition or it had completed then crashed. I suspect the former looking at the status display. The recovery partition wasn't important - it won't have changed since the backup. I have restored same workstation in the past without issue.

Booted recovery media from Ventoy USB drive. If I'm bored wrapping presents this afternoon, might try it again booting off a direct Veeam recovery drive. Also, I've got the last Veeam agent v6.2.0.121 on the Ventoy stick as well. Might try that too.

[Gostev]

I'm locking this thread now since it is becoming a place for posting random issues including of "not sure this is related to the 12.3" type

After a few weeks and with over 50K backup servers reporting to be on 12.3 already, we're not seeing any common issues in Support.
However, we will be sure to create the usual Top Issues tracker sticky thread should any surface in future.