On 11 January 2016 at 16:46, Charlie Boisseau <char...@fluency.net.uk> wrote: >> https://github.com/jwbensley/Etherate > > That looks great - definitely a useful alternative to iPerf when testing L2 > circuits. Thanks! > > The main reason we’re looking for handheld testers is really just for > simplicity’s sake - for example we have had issues in the past where not all > 1G ethernet ports on laptops seem to be capable of a true 1G (especially on > Windows which is what all our engineers use). From what I’ve seen, the > testers that the likes of Virgin Media use seem to provide a plethora of > useful results and can be operated by a monkey. The main thing is that they > #JustWork - very important with our field engineers, who are more > process-driven than technical.
Those are some of the exact same reason I started to write Etherate. Firstly because I'm trying to partition that all FEs should have Linux laptops instead - sometimes when I'm in the lab and I remove my serial to usb converter and then plug it back in again (whilst I'm booted into Windows) it won't work again until I reboot my laptop - just simple little things like that, Windows has never shown it's self to be fit for purpose in any way (I don't want to start a rant about that now!). Also if you look at libraries like Netmap and DPDK, people are pushing 10Gs using 1.6Ghz single core processors. Again, Windows == bag of shite, one should just be able to plug in your laptop and smah out 1Gbps of traffic (or 10G if you have a 10G NIC) and send traffic with a very low latency, modern day laptops are plenty fast enough, its the pony software they run. Cheers, James.
