OneDrive Backup

[Boaz]

Hello,

i've set up a new tenant with Veeam for office 365 -> doing a full backup.
im getting the below error on one of the users OneDrive Backup:

processing OneDrive *** (https://***-my.sharepoint.com/personal/***) finished with warning: Failed to backup item version: /_vti_history/512/Documents/***/OneNote_RecycleBin/OneNote_DeletedPages.one, version: 1.0, Download request retry timeout exceeded. :: 0:36:19

i tried empting the One-note recycle bin but still getting this.

any suggestions on how to resolve are welcome.
Thanks.

[Gostev]

Hello! Please include the support case ID for this issue, as requested when you click New Topic. Thanks!

[Boaz]

Sorry, newby - i'll create a support case and post back.

[Boaz]

support case number regarding "one drive backup fails" is 04711968

[Mike Resseler]

Thanks for the support case @Boaz

Please let us know what the outcome is after research of our support engineers, or if it needs to be escalated.

[tolrol]

Hi, any update on this issue? I'm experiencing this on ONLY one single user :: Processing *** (https://***-my.sharepoint.com/personal/***) finished with warning: Failed to backup item: /personal/***/Documents/*** @ ***/***.one, Download request retry timeout exceeded. :: 0:32:24 (Backup of OneNote file .ONE - OneDrive for Business. Veeam Backup for Microsoft Office 365 5.0.1.225).

ID: #02165224

[nikolaj]

Hi Rolf,

I assume that you've already tried the workaround proposed by our support engineers on May 28 and it didn't work, is that right?
The case has been reopened and is currently being investigated. So I'd recommend just staying put until a proper solution is found.

Thanks,
Nikolaj

[Lewpy]

I am seeing this problem with a customer on v5.0.2.22, Using Modern Authentication and 10 Backup Application helpers.
The timeouts are taking about 8 hours, which with 3 additional retries delays the backup considerably (it doesn't help that an glacially slow Teams backup is concurrently running at the moment too, 152hrs and counting on that).
Is there a public fix/workaround, or if we raise a case is there a fix available?

[Polina]

Hi Lewis,

The root cause may vary in this case. Please open a support case so that we first understand what type of files are processed on retries, and share your case ID here for the followup.

Thanks!

[Lewpy]

Hi Polina,

Sorry I wasn't clear in my post: the file type is also a OneNote document (.one) that someone has in their OneDrive. I don't believe it has been "correctly" uploaded to OneDrive using OneNote itself, just an old OneNote document that has moved with their Documents folder to OneDrive.
I will discuss with customer about raising a Support Case, or whether it is simpler to see if we can remove the offending file from OneDrive.

Thanks,
Lewis.

[Polina]

Hi Lewis,

If the issue is not massive and only occurs on a single file, then removing it (or re-uploading it to OneDrive in a more accurate way) would be the fastest solution indeed. Otherwise, our support engineers are always at your service and happy to help.

Thanks!

[Lewpy]

Hi Polina,

I patched the customer to the latest v5.0.2.42 release yesterday, and the timeout is still happening on the same file in the same user's OneDrive
I've looked in to this a bit more this morning, and it is a Student's OneNote document within a Class Notebook in OneDrive.
Browsing in to the Class Notebook takes a bit of URL manipulation (https://support.microsoft.com/en-us/top ... 30b52d9d00) but then allows seeing the files held inside.
The file in question is 5.59GB: would that be too big to backup? Any already known issues with Class Notebooks?
Interestingly, in the Veeam Explorer for Microsoft OneDrive for Business, I can't drill in to the Class Notebook at all, and just see it as one entity (same as the view in OneDrive without the URL manipulation). So restoring would be "all or nothing". The Explorer does know that its type is EDU Notebook, lists all the Teacher and Student access permissions, the version, but interestingly no mention of the size of it (just number of child items): guessing this is because it is actually a folder in itself?
tbh, I've not browsed OneNote notebooks before, so this may be default behaviour.

Thanks,
Lewis.

[Polina]

OneNote of 5.59GB is too big for OneDrive itself
According to Microsoft, "OneNote notebooks saved in OneDrive or SharePoint are limited to 2GB". In some cases, like yours, this limit is exceeded and this is a known unexpected behavior (reported to Microsoft) that may affect backup as well.

As for browsing OneNote notebooks in Veeam Explorers, you guess it right - it's a single entity from both backup and restore perspectives, and since it's a folder Explorers don't provide the size of items within it.

Thanks!

[Lewpy]

Hi Polina,

Is there any method to skip these files? A logic check of "OneNote file greater than 2GB -> do not attempt to download as will fail"?
As it stands, the backup is taking longer than 24 hours with retries, and that is mostly "Download request retry timeout exceeded" against this file. I have therefore had to turn retries off, to ensure that the job runs fully every 24 hours.
Having looked inside the Class Notebook, I can see other files that are even bigger (8GB), so it is not a one-off. And it's not practical to delete this file (or all of them over 2GB) as these are Student coursework submissions to their Tutors, and appear to be usable within OneNote itself.

Thanks,
Lewis.

[Polina]

Lewis,

Please involve the support team to troubleshoot. Either our engineers will offer you a nice workaround or will escalate the issue to RnD for further research.
Don't forget to post your case ID here for the follow-up.

Thanks!

[Lewpy]

Hi Polina,

I've finally raised a case for this : Case #05146022 — OneDrive backup of OneNote file over 2GB in size causes large timeout and warning
I was waiting for the initial seeding of the customer's Teams data to complete, and [honestly!] this took over 2 months to complete successfully! But now it is seeded, the incremental backup of Teams happens in an overnight time window.
The OneDrive backup is in a separate backup job running on a separate proxy server, but any support diagnosis may have required interrupting the running Teams job, so I parked the issue until now.
But to be fair, Windows Updates had to be applied, so the seeding job had to restart several times anyway, each time getting further along the process.

I understand that Veeam probably cannot backup the file using the APIs that Microsoft provide, but a workaround would be to check the file size first and if larger than 2GB then avoid even attempting to back it up? Or reduce the timeout? The apparent timeout of almost 8 hours for the affected OneDrive account seems excessive at the moment.

Thanks,
Lewis.

[Lewpy]

I've closed my Case with no improvement at this stage, and a "feature request" should have gone through to R&D about trying to mitigate this scenario.
Hopefully something is possible to implement that will either automatically skip these "oversized" files, or allow control of the timeout & retry values (which appear to be somewhere around 4 minutes timeout, and about 10 retries, based upon some very, very basic maths [so could be totally wrong]).
The good thing is I've got confidence that the rest of the user's OneDrive is being backed up, as Support analysed the log files and found that 13 files have this issue, whereas the VBO365 GUI only tells me about the first one. This makes it look like there is just a single file with a problem (whereas there are 13), and doesn't really say "but don't worry, everything else is backed up". Knowing that 13 files had the issue tells me it carries on after encountering the problem. Maybe some backup statistics per user would be useful, to show what has actually happened with that account during backup, so you can determine that everything else is safely backed up.

[Mike Resseler]

@Lewpy

Unfortunately you have closed the case already (I apologize for missing this thread). I do believe that this should be looked at by the R&D folks so I requested them to look into it anyway. It might be true that we need to skip it if it is too large, but it could also be something else and I would like to understand it if that is the case

[Lewpy]

Hi Mike,

There are some log files associated with my case, if that helps.
If you want any fresh information, or want me to try anything, reach out to me and I'll happily raise a new case.

Thanks,
Lewis.

[Mike Resseler]

Perfect @Lewpy

I cannot verify if those log files are still there (GDPR and such...) and we have a strict policy to remove log files once they are not needed anymore (again, GDPR and legislations and so on...). But I am sure the DEVs will let me know today or tomorrow if they need information and I will gladly use your offer then