Hello,
To wrap up this last issue: the 500 errors backing up to S3 B2 are network contention related. There is some kind of contention with my link to B2 and having 8 VMs backing up simultaneously sometimes triggers these errors. Last weekend I reduced concurrency to 3 and every VM was successfully backed up - no VM level retries. I can't tell if some file parts were retried since these are only logged when the AWS_MAX_ATTEMPS are exhausted and the VM backup fails.
So, these are not B2 issues and the default AWS_MAX_ATTEMPS is probably fine for most scenarios. I'll leave my at 10 following rclone defaults.
Thank you,
Pedro
