Script to randomly divide full backups into routhly equal parts.

[daznis]

Hello,


I need some help with a script. I'm not a programmer so I can't really lookup what I need. I'm currently trying to divide the load on the storage server. My idea was to divide all Veeam jobs into equal parts by VM size and assign a day when the full backup job can be performed.

[oleg.feoktistov]

Hi Darius,

I don't quite understand the use case. Can you maybe elaborate more with an example?

Thanks,
Oleg

[daznis]

Well basically split the full backup jobs into weekdays. That each day would do about equal amount of "reading" based on VM size. I don't know how to explain it in English.

[daznis]

Basically I did this, but it's not really "random". It mostly works, but I'm getting errors on Jobs that have GFS enabled.
The warning it "WARNING: Enable periodic fulls, or schedule a script to create them."

Do I need to set additional things for them ?

Code: Select all

$jobs = Get-VBRJob | ?{$_.JobType -eq "Backup"} | Sort-Object -Property info.IncludedSize 

$Week = @( "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday" )

$size = @( 0, 0, 0, 0, 0, 0, 0)

foreach ($job in $jobs) {
    
    $options = Get-VBRJobOptions -Job $job

    $JobSize = $job.info.IncludedSize / 1GB -as [INT]
    
    $min=[int]($size | measure -Minimum).Minimum

    $i = $size.IndexOf($min)

    $size.Item($i) = $size.Item($i) + $JobSize

    $options.GenerationPolicy.EnableRechek = $true
    $options.GenerationPolicy.RecheckScheduleKind = "Daily"
    $options.GenerationPolicy.RecheckDays = $Week[$i]

    Set-VBRJobOptions -job $job -Options $options
    Set-VBRJobAdvancedBackupOptions -Job $job -FullBackupScheduleKind Daily -EnableFullBackup $true -FullBackupDays $Week[$i]
}

for ($i = 0; $i -lt $Week.Length; $i++) {
    write-host 
    Write-host -NoNewline $Week[$i] "Size: " $size[$i] "GB"
} 

[david.domask]

@daznis, can I ask, you want this as a "one-off" script to randomize your full backup schedules or you want it randomized each week?

I'm a bit confused why you have the EnableRecheck section of your code -- that's for HealthCheck.

If you can elaborate what you're trying to accomplish (describe your desired business logic and/or outcome of the script), we can help more.

[daznis]

@david.domask My Idea was to divide full backups as equally as possible over the week. As most of our backups have either 7 ar 14 day retention. I though that doing a health check before the full backup would eliminate some data inconsistencies. We had some issues with synthetic backups not restoring correctly.

[david.domask]

Got it, that part is clear -- but are you running active fulls or Synthetic fulls? And do you need to adjust the schedule dynamically based on what gets added to the job or this is just something that should run occasionally as you want it?

As I get it, you have several items you're trying to solve:

1. Avoid load on the storage by spreading the full backup days out
2. Do your best to assure no data inconsistencies and also spread health-checks out to avoid load on storage

The health check (HC) isn't a problem for your script I just did not understand if it was important or not for your needs.

@daznis, maybe PM me your full description in your native tongue, I'll understand it enough I'm sure (forums are English only, so let's PM). Right now I'm not sure that a script is what you need but I'm not understanding in full yet ))