There are many free and paid applications available for common Operating Systems such as MS Windows, Linux and OS X systems that can provide varying LAN throughput performance measurement. There are also self contained hardware based solutions that do provide the most accurate LAN performance measurement / stress testing.
The software applications can be Free up to several hundred dollars.
Hardware based solutions range from a few hundred dollars to tens of thousands of dollars and are typically used by manufacturers, carriers, Tier 0 to Tier 2 ISP’s and professional IT consulting organizations.
Key things to remember when running any LAN Performance testing:
1) The is a big difference between real world Ethernet Applications and controlled Lab results.
2) Many older applications are not network aware and are not coded to take advantage of optimized physical or protocol based triggers, flags and parameters.
3) Network chipsets of the LAN jacks on equipment and the configuration of the systems or LAN devices on each end that have OS caching, RAM, packet buffering and other metrics that all play into how fast a system can respond or transmit. For example gamers like buying $300 – $400 LAN cards that are optimized for LAN performance. Most consumer electronics use cheap generic LAN chipsets from China.
4) The type of wiring used for all connection points and distance between measuring devices can make a difference. Telephone wiring has specs for maximum Megahertz frequencies supported and Category rated wiring will support much higher Megahertz and Gigahertz frequencies which translates to increased throughput/performance.
5) Environmental considerations are typically hidden and invisible noise interference that may be present in controlled Lab testing of most certainly in any real world installation. These can be the presence of surrounding high frequency communications that bleed over into your test wiring, Telco cross-connect blocks – act like antennas to pickup noise and introduce into your wiring and electro-magnetic interference causes by microwaves (Wireless gear), magnets, florescent tube ballast, etc.
We recommend using either of the following Free MS Windows based applications:
Any OS FTP program – Many Operating Systems have their own file transfer program built in and any IP based OS should have a FTP application this is ideal to use.
IPSWITCH – WS PingPro Pack – http://www.whatsupgold.com/products/ws_ping-propack/try/index.aspx
IXIA IxChariot QCheck – http://www.ixchariot.com/downloads/qcheck.html
FTP applications will require a system on each end of a physical link – one running a FTP server and the other a FTP client. It is recommended to use a large Gigabit size file for testing as IP FTP protocol will try to grab more and more bandwidth as it runs.
If using the IPSWITCH – WS PingPro Pack tool, it is important to tune the parameters of the Throughput Tab. Set the Packet Count to 100, Timeout to 500ms, Delay to 100ms. For Packet Size we recommend setting at 16383 – this is a 16MB (Max the program can handle) size.