Offload tasks, Bottleneck is source?

[BrianBuchanan]

I'm trying to understand the offload process of Scale-out Backup Repositories. Some Offload jobs are called "(SOBR Name) Offload" and some are "(Job Name) Offload". I think this is because the SOBR is configured for Copy?

My SOBR performance is a hardened Linux repository (Onetime password) and is offloading to Azure. The source repository is fast, lots of CPU, lots of RAM and 12-member RAID-10 16-TB Spinning disks; backups can process at 2-GB/s (25-Gbps network cards), our Internet is only 1-Gbps, but the "(Job Name" Offload" SOBR Tiering logs this:

10/1/2021 2:52:04 PM :: Busy: Source 81% > Proxy 3% > Network 43% > Target 45%
10/1/2021 2:52:04 PM :: Primary bottleneck: Source

The SOBR Tiering job has a processing rate of 102 MB/s, which would be the speed of our internet.

How is the Source Repository being a bottleneck?

[HannesK]

Hello,
that's because something needs to be the bottleneck

As everything is working fine, I would just let it go... bottleneck analysis is usually only interesting if one value is permanently at 99%

You can also find information about bottleneck analysis in the sticky FAQ

Best regards,
Hannes