ESXi / NFS Netapp Performance ?!

[flob]

Hi there,

i've got a question about Veeam Performance in an ESXi / NFS Netapp Infrastructure.

Veeam is installed on a physical Backup Server (16 cores) => Network Backup



1 ESXi 5.0 (mgmt connected with 1 vmnic), 1 NFS Datastore (16 configured disks, connected with 1 vmnic)


2 VMs to back up (each connected in Portgroup LAN with 1 separate vmnic)


TestVM01 (40GB full of data)
TestVM02 (40GB full of data)


when i start backupjob#1 (containing TestVM01) the performance is about 40MB/s

Log result: Load: Source 99% > Proxy 23% > Network 0% > Target 0%


ok, first i thought my infrastructure was at the end of performance, BUT!!!!


When i start backupjob#2 (containing TestVM02) at the same time when backupjob#1 is running the performance for each job ist @ 40MB/s


Does anybody have an explanation for this phenomenon? No Throtteling is configured..

Greetz
flob

[Vitaliy S.]

Hello Flob,

Could data throughput be limited on the management interface of the ESXi host? What happens if you start the 3rd job? Does it have the same performance and bottleneck stats?

Thank you!

[Andreas Neufert]

Do you have 2 physical NICs connected to your VMKernel Interface that holds the IP address that is resolved when you lookup the ESX host Name?
Looking that this NICs are 1GbE, right?

In your constellation you use the VMKernel NBD Backup mode to Transport data. Each Interface is limited to ~40% of possible full speed by VMware.

I think you can increase your backup Speed by depolying a Windows VM and add this VM as a Proxy. (Hotadd mode)