Missing OneDrive data in backup

[tm67]

Hello
#07729818
Today I was informed about an older support case, which is now resolved.
The issue was that some data in the OneDrive backup was missing, but the job did say it did run successfully.
We have a lot of other OneDrive backup jobs, and I am now wondering if this is a common or known issue?
The solution was to run "Get-VBOJob -Name "YourJobName" | Start-VBOJob -RunAsync -Full" and I am now asking myself if we need to run this regularly on all jobs / all OneDrive jobs?
And why it did not display any warning or error on the job run.
I know it's an older case, and we cannot troubleshoot this environment anymore since it now has another version installed. But I am generally asking if we should be aware of this on all other jobs we manage?
It is just not possible to check all jobs and customer OneDrives manually if all data was backed up successfully or if anything is missing when backing up thousands of different OneDrive accounts.
We need to rely on the error & warning messages.
Thanks for any comment on this.
Timo

[edh]

As far as i know this happened in old version and in lastest versions sometimes the backup do a "Full" like a synthetic full when detect those behavior and "repair" the backups

[tm67]

Hi Manuel, thanks for the information. Yeah, I have also seen some "fulls," but since I did not yet see this exact behavior, that data is really completely missing in the backup. I was a bit surprised (and a few guys in my company got a bit scared that this is happening on other jobs as well - understandable..)
So hopefully it is fixed in the current versions. And if something like this is happening again, the warnings/errors work as expected.
Is there a "best practice" on if we need to run the "-RunAsync -Full" or do we just expect that on current versions this is not necessary?

[DaStivi]

the occasional "fulls" that now gets displayed is if there is a new user for example... then this user is a full-run, others still are incremental... this was a recent change in vb365...

the manual run of "start-vbojob -full" triggers a "full" resync of the folder and content structure... hence it "reads" everything, but still only downloads missing content... there where bugs in older versions, when in sharepoint (and onedrive) users create and quickly delete and recreate files and under some edge circumstances the logic how vb365 see's new data gets corrupted and might miss some files to backup!

i can just imagine it works somewhat similiar how the file/nas backup in VBR is working... like building hashes and only dig deeper if hashes has changes.... but thats just my personal simplified explanation on how it might work.
there where also different bugs, when for example, jetDB backups got migrated to S3, data was still there but only visible through the org-restore processes and not through the job-restore or the web restore-portal ....