Maximizing File Transfer Performance Using 10Gb Ethernet and Virtualization

ManualsBrandsIntel ManualsOtherIntel 10 Gigabit CX4 Dual Port Server Adapter

Hardware

Intel® Xeon® processor X5560 series @ 2.8 GHz (8 cores, 16 threads); SMT, NUMA,

VT-x, VT-d, EIST, Turbo Enabled (default in BIOS); 24 GB Memory; Intel 10GbE

CX4 Server Adapter with VMDq

Test Methodology

RAM disk used, not disk drives. We are focused on network I/O, not disk I/O

What is being

transferred?

Directory structure, part of Linux repository: ~8 G total, ~5000 les, variable le

size, average le size ~1.6 MB

Data Collection

Tools Used

Linux * utility “sar”: Capture receive throughput and CPU utilization

Application

Tools used

Netperf (common network micro-benchmark); OpenSSH, OpenSSL (standard

Linux layers); HPN-SSH (optimized version of OpenSSH); scp, rsync (standard

Linux le transfer utilities); bbcp (“BitTorrent-like” le transfer utility)

SOURCE SERVER

NETPERF

 SCP RSYNC

SSH HPNSSH

RHEL 5.3 64-bit



DESTINATION SERVER

NETPERF

 SCP RSYNC

SSH HPNSSH

RHEL 5.3 64-bit



File Transfer

Direction



Directly connected

back-to-back

Figure 3.

FedEx and Intel delved deeper into the







featured Red Hat Enterprise Linux (RHEL)







equipped with Intel 10 Gigabit AF DA Dual

Port Server Adapters supporting direct



cable connections and the VMDq feature



Native Test Configuration

Figure 3 details the components of



systems were connected back-to-back

over 10G to eliminate variables that

could be caused by the network switches



best-case scenario because any switches



real-world scenarios, possibly degrading



A RAM disk, rather than physical disk

drives, was used in all testing to focus on

the network I/O performance rather than



The default bulk encryption used in

OpenSSH and HPN-SSH is Advanced





The application test tools included the



netperf: This commonly used

network-oriented, low-level synthetic,

micro-benchmark does very little



It is effective for evaluating the capabilities



OpenSSH, OpenSSL: These standard

Linux layers perform encryption for



HPN-SSH: This optimized version

of OpenSSH was developed by the



For more details, visit www.psc.edu/

networking/projects/hpn-ssh/

scp: The standard Linux secure copy utility

rsync: The standard Linux directory

synchronization utility

bbcp: A peer-to-peer file copy utility



the Stanford Linear Accelerator Center

