"reverse" monitoring of VBR activity

[tinto1970]

Good day all, one customer of mines experienced this incident: their primary repo (DataDomain) has been down for a whole weekend due to net issues: noone noticed it and backups stuck for almost three days.

The customer asked us to set a mail alert like "hey there's been no activity (backups of VMs, SQL logs and so on) for the last 6 hours" to be warned of similar situations. I had no idea n how and if is possible to do this with ONE. Usually I love the "protected VMs report", but this was not fitting customer's desire.

If anyone has some idea about this, I would be grateful

P.S.: me and chatgpt wrote a ps script that do the dirty job, but it works fine with regular imagebased VMs backups:
https://github.com/tinto1970/VeeamPSCol ... imojob.ps1

[jorgedlcruz]

Hello Alessandro,
We do have multiple alarms for this purpose, and you are not the first missing them as I usually conduct calls with large Enterprises that totally unaware of these. We do have specific alarms based in RPO (meaning if a restore point is not being created for the specified workloads during x time, we trigger the alarm), these ones are my favorite ones as these are beyond the noise of a job failure, these are actionable, "The VM AD-DNS do not have a restore points for the last x hours"

Full list of RPO-based alarms:

• VM with no backup (VMware vSphere)
• VM with no backup (Hyper-V)
• Computer with no backup (For Agents)
• Backup Copy RPO
• Cloud instance RPO (Since v12.3, available for all Public Cloud Hyperscalers, and all supported workloads and types of protection)
• VM Backup RPO (oVirt KVM, Proxmox VE, Nutanix AHV)
• Organization with no backup
• Application with no recent data backup sessions (For Managed Enterprise Plugins: Oracle RMAN, SAP HANA, Microsoft SQL Server since v13.0.1,, etc.)

These alarms are the absolute best, as some of them can be filtered super granularly to avoid alarm-fatigue, so you can notify on those crown jewels only that are missing the RPO, for example missing one day Warning, missing 2/3 days an Error, sending notifications to all supported paths we have in Veeam ONE (Email, Syslog, SNMP Trap, ServiceNow, and Custom script like to Teams or Slack, etc.)

I think we have this really well covered in Veeam ONE, please give it a try and let us know if something can be improved.

Thank you so much

[tinto1970]

thank you Jorge, I agree with you that ONE already has a lot of useful pre build reports. In the past I've been spending a lot of time as a (non professional) "ONE evangelist" trying to convince customers and colleagues to use it. I also wrote a blog post last year: https://blog.tinivelli.com/my-veeam-one ... c5f442899d

At the moment the customer is actually interested about being alerted when a repo is down, because they had a recent "trauma" of this kind. Their focus is not on the SLA and protected objects (as normally should be) but on the infrastructure.

This is typical in large Enterprises, where backup team do backup, infrastructure team takes care of systems, network of the net and so on. In this situation Backup team wants to recieve an alert from Veeam about a issue on the *infrastructure*. But if the Datadomain is not reacheable via network, the alert should come IMMEDIATELY from the net/infrastructure monitoring system, not from Veeam.
My answer would be "backup team: monitor SLA alerts on ONE; network team: monitor the networks; storage team: monitor the DD". But I can't be so... impolite

Would it be possible to set up ONE to send an email when one or some Repos are not available? This could be the solution in this situation.

Thank you very very much for your attention.

[jorgedlcruz]

Hello Alessandro,
Yes, we have alarms for when Backup Infrastructure components went down:


Please take a look at those

[tinto1970]

I had a deeper look at the alarm, the description reports "VBR server lost connection to the Microsoft Windows backup repository". Does it work with windows repos only?

Anyway I tested it in my lab, with a windows repo on a remote windows server: after disconnecting the vNIC the Repo appears as Unavailable in the VBR but in ONE the alert was not fired...
and in the ONE Gui I am seeing only an alert for the repo: "Backup performance data collection failure" (Failed to collect performance data for object Backup Repository 1. The RPC server is unavailable.)

[tinto1970]

edit: the alarm started working after upgrading Veeam ONE to v13
And it works for a wasabi repo too

[jorgedlcruz]

Hello Alessandro,
I remember renaming this for v13 indeed yes as the alarm is agnostic of type of repository. Thanks for checking it out, and glad it helped.