Concern about lack of reporting when volume disappears

[TitaniumCoder477]

I'm not sure if this is the best forum or the Veeam Agent one, but since the jobs are managed by BNR and job reporting is the issue, I think this is in the right place. Feel free to move it, if it's not!

Scenario

• Backup server is a VM on a VMware host

• Protected asset 1 is a server in AWS

• Protected asset 2 is a physical server on-prem

• Single Protection Group handling both assets

• Single Agent Backup job protecting both assets, sourced on the Protection Group

• Agent Backup job's Backup Mode is Volume level backup and objects are C:\, D:\, G:\, K:\, and OS volume

Today I discovered that the D:\ for protected asset 2 (the physical server) completely vanished. I only discovered this because a new tech was attempting to map a new hire to a mapped drive that used to point to it almost a decade ago. Upon further investigation, I found the RAID array had completely collapsed. Since this server is extremely old, merely functioning as a remote DC, and the data on the D:\ was completely unimportant and untouched in almost a decade, no one noticed and no one cares. So that's good.

However, that thing that has me worried is that Veeam didn't even blink an eye! That is, just now I combed through over a week of successful backups in the history and did not find where the D:\ had been protected. That means it likely disappeared months ago. And yet we've been receiving successful backup reports ever since. Is there a way to configure the job to throw a failure when a previously protected volume just disappears?

I'm sure in most cases, the end users will let us know if a production volume just disappears for a reason like this or some other reason. But we shouldn't need to rely on the canary in the coal mine! There should be some mechanism in Veeam to alert us to this, similar to how Veeam throws a warning when a previously protected asset (like a VM in a VMware backup job) is missing.

[Mildur]

Hi James

I don‘t know a method inside of veeam, which notifies you when a single disk isn‘t backed up anymore. Veeam do that only for entire vm‘s.

In my company, we are using third party monitoring systems (Paessler PRTG, there are also other linux based systems) in our environments for such cases. If a disk gets lost, we want to know that immediately and not after a few hours when the backup job runs again. In my opinion it‘s better to have monitoring system in place.

[TitaniumCoder477]

I agree in most cases losing a drive would be disastrous and probably noticeable through a number of different means. This scenario was unusual for sure. But I still do think that there should be parity and symmetry within Veeam. This has historically been one of the things I appreciated most about it. This would be the 2nd time that lack thereof has concerned me. I always ask the question of myself, "If this be so, then why not this?" For this case, if Veeam be concerned enough to warn me about an absent VM that was protected before and not now, then why is Veeam less concerned about the absence of a protected disk that was there before but not now? Parity and symmetry are evidences of architect, design, and quality control. The lack thereof is evidences of the second law of thermal dynamics at work. I see enough of that in Veeam's competitors!

[HannesK]

Hello,
I would also go for "real-time" monitoring.

I see "parity" within Veeam. We also do successful / green backup if a disk of a VM was removed. Otherwise that would probably lead to many false-positives.

From a Veeam side, the "Suspicious incremental backup size" Veeam ONE alarm might have helped.

Best regards,
Hannes

[wishr]

Hello,

Veeam ONE team here.

We've been working on a very similar request in the past. In V11a we added a new "Inclusions" column to the Workload Protection History report that highlights disks and other inclusion settings of each workload processed during every job run. This feature also supports Veeam Agent for Windows. You may schedule this report to run, let's say, daily and track the protection status of specific disks if you have specifically selected any. Please note, that if you had selected the "Entire computer" backup option, this approach won't work as we will show exactly this phrase even when a disk had been previously removed from the machine.

I would agree with Hannes that alarming is not the best solution in this case because removing disks on purpose is not an unusual thing while a failure could happen way before the scheduled backup job run.

Thanks