Backup Job Groupings

[collinp]

I haven't found a good way to group VM's within backup jobs without using powershell. If you have a single large cluster with dozens of hosts, hundreds of datastores, dozens of Petabytes and thousands of VM's, what is the best way to group VM's without code? If I choose to pin my job to a cluster, I will have a single job with thousands of VM's. If I choose by host, then as the VM's move around to different hosts, they will produce many different backup chains. if I choose by VMware datastore, datastores are removed and added weekly and we could miss backups. Tags are good, but only if the team creating the VM remembers to add the tag. Could there be a feature in the future to create backup jobs based on Dynamic groups? So we could implement something like this - an A group for machines starting with the letter A, B group for machines starting with letter B... At this point in time we have powershell scripts dynamically creating backup jobs and adding VM's to these jobs but it would be nice to get away from maintaining a script.

[collinp]

Or I could take it a step further like make a dynamic backup job that has indexing turned on for all servers larger than X terabyte. Or perform SQL backups on any vm that has *SQL* in the name. Etc.

[PetrM]

Hi Collin,

You can filter objects by name using the bottom search line in this step of the backup job creation wizard. Isn't it enough for your case, if not, may I ask you to elaborate on why it's not enough?

Thanks!

[collinp]

That step looks like it is a one time selection of VM’s to put in the job by filter. I’m looking for something where as VM’s get added to vcenter, they automatically get placed into jobs based on a filter.

[MarkBoothmaa]

What about vmtags?
https://helpcenter.veeam.com/docs/backu ... ml?ver=110
Setup jobs grouped to tags and as you deploy VM's build adding the tag into the process.

[PetrM]

Hello,

As Collin said above:

Tags are good, but only if the team creating the VM remembers to add the tag.

However, I would say that the most reliable approach is to negotiate with the team responsible for the VMs creation to always add a tag, the main purpose of the tags is to categorize objects in the virtual infrastructure. The dynamic text filter will work only if the team follows the strictly defined naming convention to create VMs, thus the likelihood that some objects are not included in the job is anyway quite high. Also, you can refer to this topic to find different examples of scripts to find VMs without tag and probably share with the team.

Thanks!