How can I back up a single vm in a large job?

[artecu]

Hi.
I am currently rolling out Veeam Enterprise 6.1 to our fleet of approx 700 vms. We have a mix of Windows and Linux and as such I have based my backups on OS and environment (using vsphere folders) eg.
Production Windows, Production Linux, QA Windows, QA LInux, Developement Windows, Development Linux, plus a couple of others for Veeam infrastructure etc.
The idea was to group similar OS backups for better dedupe results and scheduling priorities between environments ie Prod jobs are run first, then QA etc.
My Question...
Is it possible to take a quick incremental backup of a production server from a folder based backup without running the entire job or creating a seperate job for the one vm?
I followed the link below and tested but it only works when you have defined a list of vms ...sadly I dont have the knowledge to modify it to look inside the folder and exclude all vms except the one specified. Awesome script BTW.

Btw, I've suddenly come across this PowerShell trick, seems to be doing right what you were talking about.

thx Foggy
backup single VM (not all VMs) within job?

If anyone has taken this step and got it working please respond.
Many Thanks
Andy

[artecu]

OK..no takers...thats a shame... What about this then?
Can I run a script to continue a folder based backup from where it failed? If yes...can someone please let me know how?
The issue is that when a large job fails, you troubleshoot, try a fix, then you want to test the job but also make sure that the vms that were missed last time get first crack of the whip.
I would think this would be a product feature but as far as I am aware it isnt...yet!

[tsightler]

Unless I'm misunderstanding, this is actually available through the UI. If you select "retry" for a job, it will only retry the VMs that failed. I've never tried to kick that from a script though. When I get a few minutes I'll try to dig at it.

[artecu]

Thanks Tom. I have been using start! ...still learing the ropes Can you shed any light on my above request please??... in the event that I want to take a quick incremental of a server that is covered by a folder based backup job?

[tsightler]

I'm sure I could script it up if I managed to get a little time. The biggest challenge is that you have to take into account all of the possible sub-containers. I'll think about how this might be implemented. I'm thinking the simplest answer would be to simply add the folder as excluded, then add the specific VM as the object to backup, run the backup, then remove the VM and remove the folder from the excluded list. Doesn't sound difficult.

[artecu]

Great idea. I will give it a try and let you knwo.
Many thanks.

[artecu]

Hi
Because I have some disk exclusions in the job already, I didn't have the option to exclude the folder... so I removed it from the job altogether and added the target vm as you suggested. This works well. It even remembered my exclusions when I re-added the folder after the single incremental backup finished. I just tried excluding all disks from the folder to see if it would by pass all vms except the explicit one...but that is not working too well ...it has commenced with the first vm in the folder, snapshotted, processed, transferred oKb and is now onto number 2 vm etc
So the solution seems to be (and presumably could be scripted)...

1. Edit Job.
2. Take note of any exclusions you have in place (just in case veeam forgets)
3 Remove the folder/s.
4. Add the individual vm or vms explicitly.
5. Save and run the job.
6. Re-add your folder to the job and remove the explicitly defined vms, ready for your nightly backup.

This will be a good way to run a fast snap incremental backup before an upgrade, server patch etc...without the need for a new job (and associated time and space waste). A menu item for a point in time backup of any vm within a job would be a great feature for a future release IMHO.
Thanks for the idea Tom... the simplest solutions are often the best.
If someone could please post a script to do this...prompt for a jobname then a vmlist, removing and readding the folder/s afterwards it would be really appreciated.
This product is really awesome btw

[yizhar]

Hi.

Please add the following feature to Veeam:

Assuming I have a job with several VMs included.
I want to run ad-hoc (but yet incremental) backup of a single VM - backuping up only that specific VM and skipping the other.
This cannot be done in current version (only options I have are running the whole job with all servers, or running a different ad hoc full backup job).
An example of expected feature is from vmwares VDR/VDP: I can right click a single VM and "backup now" to the already existing dedup folder, using CBT etc.

Thanks,
Yizhar

[veremin]

Hi, Yizhar.

You have been merged into existing discussion; so, kindly see the answers provided above. Additionally, if VMs have been added to backup/replication jobs as individual objects, the following script would answer your requirements:

Exclude:

Code: Select all

$Jobname = Read-Host "Input job name"
$VMToBackup = Read-Host "Input VM name"
$Job = Get-VBRJob | ?{$_.name -eq $Jobname}
$Job | Get-VBRJobObject | ?{$_.name -ne $VMtobackup} | Remove-VBRJobObject 

Include VMs back:

Code: Select all

$Jobname = Read-Host "Input job name"
$Job = Get-VBRJob | ?{$_.name -eq $Jobname}
($Job | Get-VBRJobObject | ?{$_.type -eq "Exclude"}).delete() 

The algorithm regarding container exclusion is pretty good covered by Tom Sightler. However, if you have any issues with scripting it, don’t hesitate to let me know.

Thanks.

[yizhar]

Hi.

I know about the script solution which is too complex for such task.

Please consider my post as feature request, the above script is workaround for something that is missing from the product, and can be quite easily added to GUI in the future as far as I think.

Thanks,
Yizhar

[yizhar]

Hi.

And please - if possible "unmerge" my feature request or move it to the right place.
I wasn't asking for a script in that case - but suggesting feature for the GUI in next version.

Thanks,
yizhar

[veremin]

Hi, Yizhar.

Yep, I did understand that it was more like feature request, though, just tried to present you with a potential workaround that might help to bypass existing limitation.

I’ve moved this topic out from PowerShell subforum, so that it can be considered as a general feature request and not something PS specific. Meanwhile, I’ve decided not to unmerged you, since having all eggs in one basket allows us to understand better the strength of users' demand for particular feature and make a decision accordingly.

Thanks.

[yizhar]

Thanks,
Yizhar.

[AJ83]

I`m in for this feature. There shouldn`t be a need for such extensive work-around.

[cffit]

I know in past versions of VEEAM B&R if a single VM in a job failed, you had to re-run the whole job and re-backup all the VMs in the job. Has that changed in version 7 where you can just re-run the single VM that failed to backup in a job? It's kind of a pain when you have 10 VMs in a job and one fails, and then in the morning you either have to re-backup all 10 VMs or skip the backup of the single one that failed.

[foggy]

Actually, the retry operation only retries the failed VMs. There were no changes regarding that in v7, you can review the suggestions given in the topic above. Thanks.

[cffit]

Please disregard my last post. I see the retry option does just this after reading through this post. Thanks

[foggy]

No problem, glad I could point you to the right direction.

[JustinH]

I would think it would be nice to have a way to kick off a backup of a particular object in a backup without retrying all the objects in the backup job or recreating a new job.

Just an Idea

[veremin]

Hi, Justin,

How VMs are added to the backup job? As individual objects or as a part of container? In the former case, you can use the above mentioned script. In the latter, this one.

Thanks.

[domenicalvaro]

I think it would be nice to be able to open a job log and be able to right-click a VM and retry the backup just for that VM.

This would be useful when you have a job with lots of VMs and you want to just retry a few. It's would also be useful for VMs that backed up with a warning, so that we could fix the issue and try again.

Not critical, but nice to have.

Thanks,

Domenic

[veremin]

Even today you can restart just failed VMs by clicking "Retry" on job shortcut menu. Or you're after something different? Thanks.

[domenicalvaro]

Currently, you can only retry an entire job. Every failed VM in the job is backed up again, which is great in most situations.

However, I'd like to be able to retry just a single VM backup from the job, rather than the whole job. For example, if I have a job with 50 VMs and the whole job failed overnight due to some issue, I wouldn't have backup windows big enough to run the whole job again, but it would be nice to select a few critical VMs to back up manually the next morning.

Also, the ability to retry a VM backup that completed with a warning would be useful. For example, if SQL logs fail to truncate a warning is given, but the job doesn't fail. To avoid SQL logs from filling up, it would be useful to retry that VM backup again.

Actually, now that I think about it, it would probably great if you could retry any individual VM from a previous backup, including those that completed successfully.

[Vitaliy S.]

Hi Domenic,

So just to summarize your feature request - you would like to have an option to perform active full backup for any VM in the job without running the entire job, correct?

Thanks!

[dellock6]

Vitaliy, sounds more like the possibility to retry a single failed VM instead of retrying "all" the failed VMs.

[domenicalvaro]

Exactly!

[cousinsp]

We used to have lots of backup jobs with just a few VMs - now we've moved to a few backup jobs with lots of VMs, which is proving much better (faster and uses less storage). However, we recently had some problems with a few VMs - mainly Exchange servers. The transaction log was failing to truncate, which only resulted in a warning message in Veeam, so wasn't noticed for a couple of days, by which time log space on the servers was critically low.
Having fixed the underlying problem, we needed to re-run the backup. If Veeam had marked the VMs as failed, we could have just requested a 'retry', but as it was we had to rerun the whole job and back up 50 servers.
This is a request to either be able to retry backups for 'warning' VMs, or, I think more useful, to be able to retry a job for a specific selection of VMs. For example, some changes may have been made to a particular VM during the day, and it needs to be backed up.
Many thanks,
Pete Cousins.

[DGrinev]

Hi Pete,

You can use the script posted in this existing discussion or use Retry function for the failed VMs in the job.
Also, you can use VeeamZip for the particular VM to get a full backup on demand or a QuickBackup if you require to initiate an incremental run.
Please review this thread for additional information. Thank you for the feature request!

[cousinsp]

Thanks for this, the quick backup was exactly what we were looking for.
The Veeam technician who helped with our problem, doesn't know about it, I think.