Solutions to making sure no VM is missed for backup when using VM tags

[bg.ranken]

So I was wondering if there's anyone who might have a solution to a design issue we're facing.

We currently use VM tags for all of our backups jobs using a dedicated tag category called "Backup". It is working great right now and we don't want to stop using them, however the challenge we're facing is making sure we don't miss backups on any servers that aren't properly tagged.

For a while we were using VeeamONE reports in addition to the no backup alarm to keep track of this but for logistical reasons this wasn't really covering us the way we wanted to.

So our current solution is to have what we call "Unscheduled" jobs for each site. In the job we basically add that sites entire VMware cluster, then exclude any VMs that contain the Backups Tag Category. As every VM should have a tag from that category, even if it's an exclude tag, so it works out great as we only need to have someone review the job every so often and tag VMs to go in their proper scheduled job. We still have the VeeamONE "VM with no backup" alarm set to notify on VMs with no backups, but modified it to add conditions to check only powered on VMs, so we'll still get alerts if for some reason one of the tags stop working properly in a job.

But the only issue left now is that when things are working as expected and there are no untagged VMs, the jobs all throw a "Nothing to process" warning. There doesn't appear to be a way to tell Veeam that the warning is OK for only these jobs to mark the job as successful.

So I was hoping someone else might have faced a similar issue and came up with either a better solution to make sure VMs are not missed, or something to fix jobs like this were you are expecting to have no VMs in them and still want them to be marked as successful. We are trying to get all of our backup jobs to show success and this seems like the last roadblock.

[foggy]

Hi Randall, maybe some script that periodically checks whether all the existing VMs have at least one tag from that category?

[soncscy]

What about a dummy VM that's like a tiny little 10 Mb disk or something that is always untagged? It's not very...graceful I guess, but since it kind of sounds like your goal is to ensure the jobs don't throw the "nothing to process" warning, this would avoid it completely without delving into scripts, and you don't really waste and resources except a minuscule amount of datastore space. Just name the VM something informative so your team can know it's expected, or maybe edit the job description(s) for the reporting and it can help

Else, I guess you'd have to take it to scripting and just have a script that calls PowerCLI to do the same check for non-tagged VMs and if the return is null, then just don't start the job, else start the job.

[bg.ranken]

Thanks Soncscy, that seems to have done the trick.

Yes it's a bit messy, but at least that works. Wish there was a way to ignore that warning for specific jobs, but for now that achieves what I need it to.

[Gostev]

For a while we were using VeeamONE reports in addition to the no backup alarm to keep track of this but for logistical reasons this wasn't really covering us the way we wanted to.

I suppose we can ask @wishr to address this if you can share mode details with him.

[wishr]

Hi Randall,

Have you tried using the Custom Infrastructure report with a custom filter configured to exclude VMs with the "Backup" tag category? As an output, you'll get a list of VMs not having the corresponding tags assigned. You may simply schedule this report to be delivered daily and organize some process involving a manual review of the output data.

Thanks