Explanation of 'Network' in bottleneck path when no network involved

[willrussell]

Hi there,

Just a simple Q.

A client is running a backup copy job, where the source data is a RAID array of local disks and the destination data is a NAS configured on the same machine (physical, not virtual) using 1Gbps iSCSI.

In this scenario the iSCSI connection is by far the slowest, so I absolutely expect the target to be the bottleneck. In this sense, the backup is working exactly as I expect it to.

What I've noticed is the Veeam job log says -

Busy: Source 51% > Proxy 0% > Network 97% > Target 99%

So I am just trying to understand what does 'Network' mean in this scenario? Veeam surely doesn't 'know' about the iSCSI connection, as it's just presented as a block storage device?

Will

[Mildur]

Please have a lock at the FAQ, in the bottleneck analyze section:
http://forums.veeam.com/veeam-backup-re ... tml#p95291

Do you have a standalone backup server with all roles?
Or do you have different server for Proxy and Backup Repo components.

[willrussell]

Thanks Mildur, I appreciate your fast response, but the FAQ doesn't really answer my question.

I'm asking in this specific scenario - as far as Veeam should be concerned there is no network in use, so why is there a network stat at all? Surely it should be running from Source writer to Destination writer?

This is a client network. The backup server is at site 1. The offsite job in question is running on a physical host at site 2. For this job, the existing data is stored on the D: drive (local DAS on hardware RAID) and the destination is on the F: drive of the same server (iSCSI mapping over 1Gbps link).

[Mildur]

„Network in the bottleneck“ is normally the path between the data mover from the proxy server to the backup repository, in case of a backup job.

Are this components installed on the same server?
You can always get help from veeam support, if you think, that there is something wrong. Veeam support can tell you exactly what the bottleneck is in your case.

[Mildur]

This is a client network. The backup server is at site 1. The offsite job in question is running on a physical host at site 2. For this job, the existing data is stored on the D: drive (local DAS on hardware RAID) and the destination is on the F: drive of the same server (iSCSI mapping over 1Gbps link).

Then it should be on the same server. No external network between veeam components should be involved in that case. The traffic should not be leaving the backup server. I cannot say, where this 99% come from.

[willrussell]

OK, thanks for replying. There is not an problem with the job (it runs at the correct speed), but I am a VCSP like you and I just want to make sure I completely understand Veeam product behaviours as I deal with a lot of different Veeam installs!

[PetrM]

Hello,

By the way, statistics shows exactly what you expect: the bottleneck is the target - 99 %. However, if you manage to accelerate processing on the target stage, most likely the Network (data transmission between Data Movers) will become a new bottleneck.

Thanks!