Re: [WISPA] iPerf as a Real-World Performance Simulation Tool
To get a real test, you need to run iperf with numerous threads using the -P parameter. It's the only way to push 1G+ reliably. mc On Tue, Nov 12, 2013 at 12:50 PM, Sam wrote: > Has anyone used iPerf before? I'm curious of your thoughts as to the > reliability, and how the results would map to a real-world environment. > For example, a friend was testing some 3.65 radios (which had no > connection to the Internet) using iPerf. Based upon the iPerf results, a > determination was made regarding maximum concurrent users per AP and > maximum aggregate bandwidth capacity of the radio. For the maximum > concurrent users I believe they tested this at several different > download and upload speed for the fictional users. > > Having never used iPerf, I don't know if the results from their tests > (which were very poor by the way - much lower than common configurations > of DL/UL rates as well as number of users per AP I've read many of you > describing on these lists) map reasonably to what one would expect from > a given AP, or if they tend to be wonky. > > Thanks, > Sam > ___ > Wireless mailing list > Wireless@wispa.org > http://lists.wispa.org/mailman/listinfo/wireless > -- Marco C. Coelho Argon Technologies Inc. POB 875 Greenville, TX 75403-0875 903-455-5036 ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless
Re: [WISPA] iPerf as a Real-World Performance Simulation Tool
Note on IPerf! It tests and reports UPLOAD by default, not download. Check the commandline options carefully to differentiate between UDP, TCP, Up, Down and concurrency. It will gladly send 1Gbps UDP and report that it is doing so... but that doesn't mean jack! -d -r -b On Tue, Nov 12, 2013 at 8:50 PM, Sam wrote: > Has anyone used iPerf before? I'm curious of your thoughts as to the > reliability, and how the results would map to a real-world environment. > For example, a friend was testing some 3.65 radios (which had no > connection to the Internet) using iPerf. Based upon the iPerf results, a > determination was made regarding maximum concurrent users per AP and > maximum aggregate bandwidth capacity of the radio. For the maximum > concurrent users I believe they tested this at several different > download and upload speed for the fictional users. > > Having never used iPerf, I don't know if the results from their tests > (which were very poor by the way - much lower than common configurations > of DL/UL rates as well as number of users per AP I've read many of you > describing on these lists) map reasonably to what one would expect from > a given AP, or if they tend to be wonky. > > Thanks, > Sam > ___ > Wireless mailing list > Wireless@wispa.org > http://lists.wispa.org/mailman/listinfo/wireless > ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless
Re: [WISPA] iPerf as a Real-World Performance Simulation Tool
Don't remember. Solved the problem with matching machines Sent from my Verizon Wireless 4G LTE Smartphone Original message From: Josh Luthman Date: 11/12/2013 9:10 PM (GMT-05:00) To: WISPA General List Subject: Re: [WISPA] iPerf as a Real-World Performance Simulation Tool Don't use cheap chipsets on the laptop. Who shipped the laptops you had problems with? Josh Luthman Office: 937-552-2340 Direct: 937-552-2343 1100 Wayne St Suite 1337 Troy, OH 45373 On Nov 12, 2013 9:09 PM, "Bob Moldashel" wrote: We have seen download/upload irregularities when using it with different types of laptops/processors. Overall it is a good tool. Something to check first is to test laptop to laptop with a cossover cable and see what your results are first. Nothing like chasing a problem that isn't there. -B- On Nov 12, 2013, at 2:53 PM, Bret Clark wrote: I've used it on a PI and found it to be very accurate. Next to Iperf, put/get files from an FTP server provides good throughput measurements. On 11/12/2013 02:49 PM, Clay Stewart wrote: Is it accurate using it on a Raspberry-pi? I was loading up my first one today to try out Iperf. Since someone on this thread mentioned 'CPU', I was wondering if that would be an issue on a PI? Clocked at 800Mhz.. On Tue, Nov 12, 2013 at 2:35 PM, Mike Hammett wrote: iPerf is very accurate in what it tells you. What you draw from those conclusions may or may not be accurate. - Mike Hammett Intelligent Computing Solutions http://www.ics-il.com From: "Sam" To: wireless@wispa.org Sent: Tuesday, November 12, 2013 12:50:51 PM Subject: [WISPA] iPerf as a Real-World Performance Simulation Tool Has anyone used iPerf before? I'm curious of your thoughts as to the reliability, and how the results would map to a real-world environment. For example, a friend was testing some 3.65 radios (which had no connection to the Internet) using iPerf. Based upon the iPerf results, a determination was made regarding maximum concurrent users per AP and maximum aggregate bandwidth capacity of the radio. For the maximum concurrent users I believe they tested this at several different download and upload speed for the fictional users. Having never used iPerf, I don't know if the results from their tests (which were very poor by the way - much lower than common configurations of DL/UL rates as well as number of users per AP I've read many of you describing on these lists) map reasonably to what one would expect from a given AP, or if they tend to be wonky. Thanks, Sam ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless -- -- SCS Clay Stewart CEO, Tye River Farms, Inc., DBA Stewart Computer Services 434.263.6363 O 434.942.6510 C cstew...@stewartcomputerservices.com “We Keep You Up and Running” Wireless Broadband Programming Network Services ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless
Re: [WISPA] iPerf as a Real-World Performance Simulation Tool
On 11/12/2013 09:15 PM, Andrew Jones wrote: > Careful using FTP as, particularly on fast links, you may find yourself > measuring disk IO instead. > > If you can exceed modern day disk I/O on a wireless network...I'm impressed. Bret ___ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless