Slow Teams and public sharepoint job

[Johan M]

Hi all
We are currently in the process of implementing a new customer, with roughly 17500 users and 1 PB data, divided into 33 jobs, divided equally between the proxies.
The environments consists of 1 VBO server (8 core 32 GB memory), and 8 proxyes (16 cores 32 GB memory) and we are running version 6.0.0.367 of M365 backup.
We have divided the backup into 32 jobs handling the users mailboxes, archive, onedrive and personal sharepoint, based on security groups.
One job is configured to handle Teams and shared sharepoint sites runs in the same job, and this job is causing us problems, as it contains:
142799 objects and it takes roughly 70 hours to enumerate the ojects.
The job has run for 70+ hours and only went through 4500 objects out of 142799 and around 60 GB of data, we currently do not know the total amount of data. We have 50 auxiliary backup applications added to the Backup Applications Manager to mitigate/avoid throttling. We dont see any throttling in the logs for the job.
Does any of you have an idea to what we can do to speed this up? Ideally we would like this to be done within 24 hours, and first thing we will do now is to split the jobs in to two seperate jobs one performing Teams backup and the other SharePoint backup

We do not have a active case with Veeam.

[jorgedlcruz]

Thank you very much for such detailed information, we would love to help you a bit more and understand where the bottleneck can be, but first, we will need a support ticket, please.

I will definitely monitor it closely, as our Support Engineers are brilliant, and they have the proper tools to diagnose where the problem might be.

Splitting the Jobs would always help as well.

Please share the Support ticket number as soon as you have it.

Thank you!

[Polina]

Hi Johan,

I'd suggest you split the job and make it 20K object per job maximum. Note also, that each team can be considered as 3 objects in one, because from the backup perspective it consists of a mailbox, a site and teams metadata.
Here you can find some insights on the recommended sizing per job and per proxy: https://bp.veeam.com/vbo/guide/design/maxconfig.html

Thanks!